diff --git a/data/tuto1/vtt/tuto1-activite1-vid1-en.vtt b/data/tuto1/vtt/tuto1-activite1-vid1-en.vtt
index a545f086b757ca11e4ca7124b30e18a5de60a0a4..1b4c9274aa475475bae458ef85fade425bc95782 100644
--- a/data/tuto1/vtt/tuto1-activite1-vid1-en.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid1-en.vtt
@@ -1,40 +1,62 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.333 --> 00:00:02.914
+1
+00:00:00.048 --> 00:00:02.715
 What is artificial intelligence?
 
-00:00:03.541 --> 00:00:06.666
-AI, or artificial intelligence,
-is a collection of IT tools
+2
+00:00:03.381 --> 00:00:06.756
+AI or artificial intelligence
+is a set of computer tools
 
-00:00:06.750 --> 00:00:10.083
-that mimic actions
-previously only done by humans.
+3
+00:00:06.798 --> 00:00:10.506
+that mimic activities
+previously performed only by humans.
 
-00:00:10.370 --> 00:00:12.515
-They are most often cognitive actions.
+4
+00:00:10.548 --> 00:00:12.381
+These are usually cognitive activities.
 
-00:00:12.583 --> 00:00:16.084
-For example, recognising what is
-in an image or even translating a text.
+5
+00:00:12.423 --> 00:00:16.604
+For example, it recognises
+what is in a picture or translates text.
 
-00:00:16.509 --> 00:00:18.020
-We can find it all around us!
+6
+00:00:16.643 --> 00:00:18.179
+It is all around us.
 
-00:00:18.083 --> 00:00:20.729
-Our phone recognising faces on our photos,
+7
+00:00:18.215 --> 00:00:20.881
+Our phones recognise faces in our photos,
 
-00:00:20.790 --> 00:00:24.174
-translation sites that immediately
-recognise the language being spoken.
+8
+00:00:20.923 --> 00:00:22.756
+translation sites recognise the language
 
-00:00:24.521 --> 00:00:29.382
-Social media apps that can recognise faces
-on our photos so we can add rabbit ears.
+9
+00:00:22.798 --> 00:00:24.256
+we speak before we even tell them,
 
-00:00:29.458 --> 00:00:32.179
-Voice assistants
-who we can speak to. Great, right?
+10
+00:00:24.298 --> 00:00:27.756
+social networks can recognise faces
+in our pictures
+
+11
+00:00:27.798 --> 00:00:29.506
+and add bunny ears to them,
+
+12
+00:00:29.548 --> 00:00:31.131
+and we have voice assistants to talk to.
+
+13
+00:00:31.173 --> 00:00:32.131
+How cool is that?
+
+14
+00:00:32.173 --> 00:00:33.715
+Let's take a closer look.
 
-00:00:32.497 --> 00:00:33.685
-Let's take a closer look.
\ No newline at end of file
diff --git a/data/tuto1/vtt/tuto1-activite1-vid1-it.vtt b/data/tuto1/vtt/tuto1-activite1-vid1-it.vtt
old mode 100755
new mode 100644
index ccfa12c1efecd083a1cbb16fad2e4b6a61ac4727..2d3b1901075496baa0e7c3ac057b22de6cc6f9ab
--- a/data/tuto1/vtt/tuto1-activite1-vid1-it.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid1-it.vtt
@@ -1,40 +1,68 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.333 --> 00:00:02.914
-Cos'è l'intelligenza artificiale?
+1
+00:00:00.473 --> 00:00:02.806
+CHE COS'È L'INTELLIGENZA ARTIFICIALE?
 
-00:00:03.541 --> 00:00:06.666
-AI, o intelligenza artificiale,
-è un insieme di strumenti IT
+2
+00:00:02.831 --> 00:00:05.060
+L'IA o intelligenza artificiale
 
-00:00:06.750 --> 00:00:10.083
-che imitano azioni
-precedentemente fatte solo dagli umani.
+3
+00:00:05.085 --> 00:00:06.957
+è un insieme di strumenti informatici
 
-00:00:10.370 --> 00:00:12.515
-Sono più spesso azioni cognitive.
+4
+00:00:06.982 --> 00:00:10.476
+che simulano attività
+finora realizzate unicamente dagli umani.
 
-00:00:12.583 --> 00:00:16.084
-Per esempio, riconoscere ciò che è
-in un'immagine o anche tradurre un testo.
+5
+00:00:10.501 --> 00:00:12.536
+Interessa perlopiù le attività cognitive.
 
-00:00:16.509 --> 00:00:18.020
-Possiamo trovarlo tutto intorno a noi!
+6
+00:00:12.561 --> 00:00:15.290
+Ad esempio, riconoscere un'immagine
+o tradurre un testo.
 
-00:00:18.083 --> 00:00:20.729
-Il nostro telefono riconosce i volti sulle nostre foto,
+7
+00:00:15.315 --> 00:00:16.790
+La troviamo ovunque intorno a noi:
 
-00:00:20.790 --> 00:00:24.174
-siti di traduzione che riconoscono immediatamente
-riconoscono la lingua parlata.
+8
+00:00:16.815 --> 00:00:19.650
+i nostri telefoni, che riconoscono
+il nostro viso nelle foto,
 
-00:00:24.521 --> 00:00:29.382
-App per i social media in grado di riconoscere i volti
-sulle nostre foto così possiamo aggiungere orecchie da coniglio.
+9
+00:00:19.681 --> 00:00:22.400
+i siti di traduzione
+che riconoscono la lingua parlata
 
-00:00:29.458 --> 00:00:32.179
-Assistenti vocali
-con cui possiamo parlare. Fantastico, vero?
+10
+00:00:22.425 --> 00:00:23.994
+ancor prima di averla impostata,
+
+11
+00:00:24.019 --> 00:00:26.080
+i social network, capaci di riconoscere
+
+12
+00:00:26.105 --> 00:00:28.966
+il nostro viso nelle foto
+per aggiungere delle orecchie di coniglio,
+
+13
+00:00:28.991 --> 00:00:31.234
+gli assistenti vocali,
+con cui possiamo parlare.
+
+14
+00:00:31.259 --> 00:00:32.360
+Straordinario, no?
+
+15
+00:00:32.385 --> 00:00:33.879
+Diamo un'occhiata più da vicino.
 
-00:00:32.497 --> 00:00:33.685
-Diamo un'occhiata più da vicino.
\ No newline at end of file
diff --git a/data/tuto1/vtt/tuto1-activite1-vid1-sl.vtt b/data/tuto1/vtt/tuto1-activite1-vid1-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..a03727ee2c115225eae8545aead1eefdc1dbec20
--- /dev/null
+++ b/data/tuto1/vtt/tuto1-activite1-vid1-sl.vtt
@@ -0,0 +1,65 @@
+WEBVTT FILE
+
+1
+00:00:00.048 --> 00:00:02.715
+Kaj je umetna inteligenca?
+
+2
+00:00:02.982 --> 00:00:06.514
+Umetna inteligenca (UI)
+je sklop računalniških orodij,
+
+3
+00:00:06.624 --> 00:00:08.271
+ki posnemajo aktivnosti,
+
+4
+00:00:08.359 --> 00:00:10.524
+ki smo jih doslej povezovali le
+s človeškim umom.
+
+5
+00:00:10.549 --> 00:00:12.572
+Ponavadi gre za kognitivne aktivnosti.
+
+6
+00:00:12.669 --> 00:00:16.437
+UI lahko recimo prepozna,
+kaj je na sliki, ali prevede besedilo.
+
+7
+00:00:16.656 --> 00:00:18.289
+UI je povsod okrog nas.
+
+8
+00:00:18.383 --> 00:00:20.540
+Telefoni prepoznajo obraze
+na podobah,
+
+9
+00:00:20.565 --> 00:00:24.265
+spletni prevajalniki samodejno
+prepoznajo jezik naše komunikacije.
+
+10
+00:00:24.361 --> 00:00:27.773
+Družbena omrežja
+prepoznajo obraze na naših fotografijah
+
+11
+00:00:27.798 --> 00:00:29.506
+in jim dodajo zajčja ušesa.
+
+12
+00:00:29.548 --> 00:00:31.539
+Govorimo lahko
+z virtualnimi asistenti.
+
+13
+00:00:31.587 --> 00:00:32.545
+A ni to super?
+
+14
+00:00:32.570 --> 00:00:33.715
+Poglejmo pobliže.
+
diff --git a/data/tuto1/vtt/tuto1-activite1-vid2-en.vtt b/data/tuto1/vtt/tuto1-activite1-vid2-en.vtt
index 5ba8ef06222273b20d947581a2edced18d433e1d..96be5c505651dbc956f53b1ef6b8ee6623ae6382 100644
--- a/data/tuto1/vtt/tuto1-activite1-vid2-en.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid2-en.vtt
@@ -1,39 +1,53 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.167 --> 00:00:01.250
+1
+00:00:00.090 --> 00:00:01.131
 Magic?
 
-00:00:02.312 --> 00:00:05.660
-For a program to learn to recognise
-images, you have to train it first.
+2
+00:00:01.881 --> 00:00:06.335
+The first step in making a program learn
+to recognise images is to train it.
 
-00:00:06.014 --> 00:00:09.330
-This is what is known
-as "machine learning."
+3
+00:00:06.607 --> 00:00:09.940
+This is what we call machine learning.
 
-00:00:09.992 --> 00:00:12.250
-We'll show it examples,
-and tell it what they are.
+4
+00:00:09.965 --> 00:00:12.552
+We'll show examples
+by telling it what it is.
 
-00:00:12.640 --> 00:00:15.312
-This is an image of a cat.
-This is an image of a dog.
+5
+00:00:12.590 --> 00:00:15.465
+Whether it's a picture of a cat or a dog,
 
-00:00:15.375 --> 00:00:18.625
-We'll show it thousands
-for each thing we want it to recognise.
+6
+00:00:15.506 --> 00:00:17.173
+we'll show it thousands of examples
 
-00:00:19.167 --> 00:00:21.417
-At the start it will make
-a lot of mistakes.
+7
+00:00:17.215 --> 00:00:18.506
+of each thing we want it to recognise.
 
-00:00:21.458 --> 00:00:24.916
-But after seeing plenty of examples,
-it will recognise an image of a cat,
+8
+00:00:19.173 --> 00:00:21.340
+At first, it will make many mistakes,
 
-00:00:24.937 --> 00:00:26.958
-even on a photo that it has never seen.
+9
+00:00:21.381 --> 00:00:23.256
+but after seeing numerous examples,
+
+10
+00:00:23.298 --> 00:00:25.048
+it will be able to identify
+a picture of a cat
+
+11
+00:00:25.090 --> 00:00:26.965
+even in a photo it has never seen.
+
+12
+00:00:27.006 --> 00:00:29.548
+Now we can train our own AI.
 
-00:00:27.042 --> 00:00:29.583
-We're now going
-to be able to train our own AI!
\ No newline at end of file
diff --git a/data/tuto1/vtt/tuto1-activite1-vid2-it.vtt b/data/tuto1/vtt/tuto1-activite1-vid2-it.vtt
old mode 100755
new mode 100644
index 920ac40059bf376ad9775acb9f8d343ff0e93a01..63183df65fa6a29450527765eb9bbf46e9233786
--- a/data/tuto1/vtt/tuto1-activite1-vid2-it.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid2-it.vtt
@@ -1,39 +1,60 @@
-WEBVTT
-
-00:00:00.167 --> 00:00:01.250
-Magia?
-
-00:00:02.312 --> 00:00:05.660
-Per un programma che impara a riconoscere
-le immagini, devi prima addestrarlo.
-
-00:00:06.014 --> 00:00:09.330
-Questo è ciò che è noto
-come "apprendimento automatico".
-
-00:00:09.992 --> 00:00:12.250
-Gli mostreremo degli esempi,
-e gli diremo cosa sono.
+WEBVTT FILE
+
+1
+00:00:00.560 --> 00:00:01.856
+MAGIA?
+
+2
+00:00:01.881 --> 00:00:04.353
+Perché un programma impari
+a riconoscere delle immagini,
+
+3
+00:00:04.378 --> 00:00:05.633
+è necessario addestrarlo.
+
+4
+00:00:05.658 --> 00:00:08.048
+È ciò che chiamiamo
+"apprendimento automatico"
+
+5
+00:00:08.073 --> 00:00:09.940
+o "<i>machine learning</i>", in inglese.
+
+6
+00:00:09.965 --> 00:00:13.321
+Gli vengono mostrati degli esempi,
+specificando di cosa si tratta:
+
+7
+00:00:13.346 --> 00:00:16.201
+questa è l'immagine di un gatto,
+questa è l'immagine di un cane...
+
+8
+00:00:16.226 --> 00:00:19.676
+Gliene vengono mostrati migliaia
+per ogni cosa che vogliamo che riconosca.
 
-00:00:12.640 --> 00:00:15.312
-Questa è un'immagine di un gatto.
-Questa è l'immagine di un cane.
+9
+00:00:19.735 --> 00:00:21.582
+All'inizio, sbaglierà, e tanto,
 
-00:00:15.375 --> 00:00:18.625
-Gliene mostreremo migliaia
-per ogni cosa che vogliamo che riconosca.
+10
+00:00:21.655 --> 00:00:23.256
+ma dopo aver visto molti esempi,
 
-00:00:19.167 --> 00:00:21.417
-All'inizio farà
-molti errori.
+11
+00:00:23.281 --> 00:00:25.557
+sarà in grado di riconoscere
+l'immagine di un gatto
 
-00:00:21.458 --> 00:00:24.916
-Ma dopo aver visto molti esempi,
-riconoscerà l'immagine di un gatto,
+12
+00:00:25.582 --> 00:00:27.241
+anche in una foto che non ha mai visto.
 
-00:00:24.937 --> 00:00:26.958
-anche su una foto che non ha mai visto.
+13
+00:00:27.266 --> 00:00:29.813
+Adesso, possiamo addestrare la nostra IA.
 
-00:00:27.042 --> 00:00:29.583
-Ora saremo in grado
-essere in grado di addestrare la nostra IA!
\ No newline at end of file
diff --git a/data/tuto1/vtt/tuto1-activite1-vid2-sl.vtt b/data/tuto1/vtt/tuto1-activite1-vid2-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..27369c6e48ff241c459ae675e8b21ec6ec969956
--- /dev/null
+++ b/data/tuto1/vtt/tuto1-activite1-vid2-sl.vtt
@@ -0,0 +1,58 @@
+WEBVTT FILE
+
+1
+00:00:00.191 --> 00:00:01.232
+Čarovnija?
+
+2
+00:00:01.881 --> 00:00:04.381
+Če želimo,
+da program prepozna določeno podobo,
+
+3
+00:00:04.406 --> 00:00:06.335
+ga moramo tega najprej naučiti.
+
+4
+00:00:06.430 --> 00:00:08.195
+Temu pravimo strojno učenje.
+
+5
+00:00:08.220 --> 00:00:11.759
+Pokažemo mu primere
+in za vsakega povemo, kaj predstavlja.
+
+6
+00:00:11.829 --> 00:00:13.907
+Naj gre za podobo mačke ali psa,
+
+7
+00:00:13.946 --> 00:00:16.727
+vsakič mu pokažemo
+na tisoče primerov tistega,
+
+8
+00:00:16.794 --> 00:00:18.833
+kar hočemo, da se nauči prepoznati.
+
+9
+00:00:18.858 --> 00:00:20.716
+Na začetku se bo velikokrat zmotil,
+
+10
+00:00:20.741 --> 00:00:22.724
+ko pa si bo ogledal veliko
+število primerov,
+
+11
+00:00:22.774 --> 00:00:24.687
+bo že znal prepoznati podobo mačke,
+
+12
+00:00:24.757 --> 00:00:26.981
+tudi na sliki, ki je ni še nikdar videl.
+
+13
+00:00:27.044 --> 00:00:29.586
+Začnemo lahko z učenjem svoje lastne UI.
+
diff --git a/data/tuto1/vtt/tuto1-activite1-vid3-en.vtt b/data/tuto1/vtt/tuto1-activite1-vid3-en.vtt
index 9712438c7a83f3ac56a5654d6f6169baadaab0c2..515827811d745178632e42693be849985e4933a5 100644
--- a/data/tuto1/vtt/tuto1-activite1-vid3-en.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid3-en.vtt
@@ -1,47 +1,72 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.000 --> 00:00:02.225
-We just trained our first program!
+1
+00:00:00.048 --> 00:00:02.506
+We have just trained our first programme.
 
-00:00:02.708 --> 00:00:05.833
-Well done!
-As you could see, it wasn't magic.
+2
+00:00:02.548 --> 00:00:03.548
+Well done.
 
-00:00:06.208 --> 00:00:09.751
-First, we told it what we were showing it.
-Then we showed it examples of images.
+3
+00:00:03.708 --> 00:00:06.511
+As we saw,
+there's actually nothing magical about it.
 
-00:00:10.000 --> 00:00:11.458
-That's the learning phase.
+4
+00:00:06.536 --> 00:00:08.389
+First, we told it what we would show it,
 
-00:00:11.789 --> 00:00:13.854
-We call the program
-we trained "the model".
+5
+00:00:08.414 --> 00:00:10.586
+and then we showed it some example images.
 
-00:00:14.610 --> 00:00:17.790
+6
+00:00:10.633 --> 00:00:11.998
+This is the learning phase.
+
+7
+00:00:12.023 --> 00:00:14.497
+The programme being trained
+is called a 'model'.
+
+8
+00:00:14.857 --> 00:00:18.065
 It can now predict
-the category an image belongs to.
+which category an image belongs to.
 
-00:00:18.242 --> 00:00:22.040
-Its prediction is still a statistic
-approximation, hence the percentage.
+9
+00:00:18.090 --> 00:00:20.798
+Its prediction
+is still a statistical approximation,
 
-00:00:22.458 --> 00:00:24.625
+10
+00:00:20.840 --> 00:00:22.131
+hence the percentage.
+
+11
+00:00:22.340 --> 00:00:24.915
 According to its calculation,
-there's more chance
+there is a greater chance
 
-00:00:24.687 --> 00:00:27.208
+12
+00:00:24.940 --> 00:00:27.690
 that this image belongs
-to one category over another.
+to this category than another.
+
+13
+00:00:27.966 --> 00:00:29.590
+But can it detect everything?
 
-00:00:27.994 --> 00:00:29.625
-But can it recognise everything?
+14
+00:00:29.631 --> 00:00:32.173
+If we teach it to recognise cats,
 
-00:00:29.917 --> 00:00:32.333
-If we teach it to recognise cats or dogs,
+15
+00:00:32.215 --> 00:00:34.423
+could it identify lions or tigers?
 
-00:00:32.416 --> 00:00:34.479
-can it recognise lions or tigers?
+16
+00:00:34.465 --> 00:00:36.090
+Can our programme be wrong?
 
-00:00:34.541 --> 00:00:36.083
-Can our program get it wrong?
\ No newline at end of file
diff --git a/data/tuto1/vtt/tuto1-activite1-vid3-it.vtt b/data/tuto1/vtt/tuto1-activite1-vid3-it.vtt
old mode 100755
new mode 100644
index fb73d9a2b9e39716f88ec22cb9879c4164ce1a05..7fc8cf2f0541a9d8c61c18df601eedcfd9416f99
--- a/data/tuto1/vtt/tuto1-activite1-vid3-it.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid3-it.vtt
@@ -1,47 +1,76 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.000 --> 00:00:02.225
-Abbiamo appena addestrato il nostro primo programma!
+1
+00:00:00.020 --> 00:00:02.526
+ABBIAMO APPENA ADDESTRATO
+IL NOSTRO PRIMO PROGRAMMA!
 
-00:00:02.708 --> 00:00:05.833
-Ben fatto!
-Come avete potuto vedere, non è stata una magia.
+2
+00:00:02.551 --> 00:00:03.560
+Complimenti!
 
-00:00:06.208 --> 00:00:09.751
-Per prima cosa, gli abbiamo detto cosa gli stavamo mostrando.
-Poi gli abbiamo mostrato esempi di immagini.
+3
+00:00:03.585 --> 00:00:05.980
+Come abbiamo potuto vedere,
+non c'è nulla di magico.
 
-00:00:10.000 --> 00:00:11.458
+4
+00:00:06.051 --> 00:00:08.428
+Prima, gli abbiamo detto
+cosa gli avremmo mostrato
+
+5
+00:00:08.453 --> 00:00:10.722
+poi, gli abbiamo mostrato
+immagini di esempio.
+
+6
+00:00:10.747 --> 00:00:12.414
 Questa è la fase di apprendimento.
 
-00:00:11.789 --> 00:00:13.854
-Chiamiamo il programma
-che abbiamo addestrato "il modello".
+7
+00:00:12.439 --> 00:00:15.023
+Chiamiamo "modello"
+il programma che addestriamo.
 
-00:00:14.610 --> 00:00:17.790
-Ora può predire
-la categoria a cui appartiene un'immagine.
+8
+00:00:15.048 --> 00:00:18.382
+Adesso è in grado di indovinare
+a quale categoria appartiene un'immagine.
 
-00:00:18.242 --> 00:00:22.040
-La sua previsione è ancora una statistica
-approssimazione, da cui la percentuale.
+9
+00:00:18.407 --> 00:00:20.941
+Le sue previsioni
+restano approssimazioni statistiche,
 
-00:00:22.458 --> 00:00:24.625
-Secondo il suo calcolo,
-c'è più possibilità
+10
+00:00:20.966 --> 00:00:22.435
+da qui le percentuali.
 
-00:00:24.687 --> 00:00:27.208
-che questa immagine appartenga
-ad una categoria piuttosto che ad un'altra.
+11
+00:00:22.460 --> 00:00:24.977
+Secondo i suoi calcoli,
+ci sono più probabilità
 
-00:00:27.994 --> 00:00:29.625
+12
+00:00:25.002 --> 00:00:28.230
+che quell'immagine appartenga
+a tale categoria piuttosto che a un'altra.
+
+13
+00:00:28.255 --> 00:00:29.669
 Ma può riconoscere tutto?
 
-00:00:29.917 --> 00:00:32.333
-Se gli insegniamo a riconoscere i gatti o i cani,
+14
+00:00:29.769 --> 00:00:32.298
+Se lo addestriamo
+a riconoscere dei cani o dei gatti,
+
+15
+00:00:32.323 --> 00:00:34.448
+potrà riconoscere dei leoni o delle tigri?
 
-00:00:32.416 --> 00:00:34.479
-può riconoscere i leoni o le tigri?
+16
+00:00:34.473 --> 00:00:36.506
+Il nostro programma potrebbe sbagliarsi?
 
-00:00:34.541 --> 00:00:36.083
-Può il nostro programma sbagliare?
diff --git a/data/tuto1/vtt/tuto1-activite1-vid3-sl.vtt b/data/tuto1/vtt/tuto1-activite1-vid3-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..0ed20c7437bd83e1cb0798d52a8fcc734a8d78f1
--- /dev/null
+++ b/data/tuto1/vtt/tuto1-activite1-vid3-sl.vtt
@@ -0,0 +1,58 @@
+WEBVTT FILE
+
+WEBVTT
+
+00:00:00.048 --> 00:00:02.506
+Uspešno smo usposobili svoj prvi program.
+
+00:00:02.548 --> 00:00:03.548
+Dobro opravljeno!
+
+00:00:03.840 --> 00:00:05.923
+Kot smo videli,
+ne gre za nobeno čarovnijo.
+
+00:00:05.965 --> 00:00:08.006
+Najprej smo programu povedali,
+kaj ga boom naučili,
+
+00:00:08.048 --> 00:00:09.673
+nato pa smo mu pokazali veliko število primerov.
+
+00:00:09.881 --> 00:00:11.590
+To je faza učenja.
+
+00:00:11.631 --> 00:00:13.840
+Takšnemu učečemu se programu pravimo 'model'.
+
+00:00:14.423 --> 00:00:17.631
+Zdaj zna že predvideti,
+v katero kategorijo sodi določena podoba.
+
+00:00:18.090 --> 00:00:20.798
+Njegovo predvidevanje
+je sicer še vedno statistični približek,
+
+00:00:20.840 --> 00:00:22.131
+izražen v odstotkih.
+
+00:00:22.340 --> 00:00:24.590
+Glede na njegove izračune
+obstaja večja verjetnost,
+
+00:00:24.631 --> 00:00:27.381
+da ta slika sodi v to kategorijo
+kot pa v kakšno drugo.
+
+00:00:27.715 --> 00:00:29.590
+Toda ali model res zazna vse?
+
+00:00:29.631 --> 00:00:32.173
+Če ga naučimo prepoznati mačke,
+
+00:00:32.215 --> 00:00:34.423
+ali bo znal prepoznati tudi leve in tigre?
+
+00:00:34.465 --> 00:00:36.090
+Se lahko naš program zmoti?
+
diff --git a/data/tuto1/vtt/tuto1-activite1-vid4-en.vtt b/data/tuto1/vtt/tuto1-activite1-vid4-en.vtt
index 80fb7353ff836030dacda553a4a1a20ebf04162e..adeae5a1fc1cb1d82291670db3d81becd02b1146 100644
--- a/data/tuto1/vtt/tuto1-activite1-vid4-en.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid4-en.vtt
@@ -1,36 +1,64 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.521 --> 00:00:04.312
-Our program can only do
-what it has been trained to do.
+1
+00:00:00.173 --> 00:00:04.465
+Our programme can only do
+what it has been trained to do!
 
-00:00:04.708 --> 00:00:07.208
+2
+00:00:04.506 --> 00:00:07.256
 The program can only recognise
-what we've shown it.
+what it has been shown.
 
-00:00:07.527 --> 00:00:11.390
+3
+00:00:07.298 --> 00:00:09.453
 If we teach it to recognise cats or dogs,
-it won't be able to recognise lions.
 
-00:00:11.458 --> 00:00:14.416
-But it can tell us
-which category it is closer to.
+4
+00:00:09.478 --> 00:00:11.506
+it won't be able to recognise lions,
 
-00:00:14.500 --> 00:00:18.042
-Cats? It is only as smart
-as we taught it to be.
+5
+00:00:11.548 --> 00:00:14.465
+but it will be able to tell us
+which categorie it is most like,
 
-00:00:18.126 --> 00:00:21.226
-When we discover something new,
-we don't know what it is straight away.
+6
+00:00:14.506 --> 00:00:15.590
+in this case, cats.
 
-00:00:21.672 --> 00:00:26.160
-But once we learn, we can recognise it
+7
+00:00:16.131 --> 00:00:18.006
+It's not more intelligent
+than what we teach it.
+
+8
+00:00:18.048 --> 00:00:19.506
+When we discover a new thing,
+
+9
+00:00:19.590 --> 00:00:21.465
+we don't know what it is at first,
+
+10
+00:00:21.506 --> 00:00:22.965
+but once we have learned it,
+
+11
+00:00:23.040 --> 00:00:26.039
+we are able to recognise it
 in different positions or contexts.
 
-00:00:26.632 --> 00:00:28.250
-Is it the same for our program?
+12
+00:00:26.124 --> 00:00:27.742
+Is it the same for our programme?
 
-00:00:28.792 --> 00:00:32.390
+13
+00:00:27.804 --> 00:00:29.749
 If we teach it to recognise lions,
-will it recognise soft toy lions?
\ No newline at end of file
+
+14
+00:00:29.774 --> 00:00:32.715
+will it be able
+to recognise cuddly toy lions?
+
diff --git a/data/tuto1/vtt/tuto1-activite1-vid4-it.vtt b/data/tuto1/vtt/tuto1-activite1-vid4-it.vtt
old mode 100755
new mode 100644
index 49617f60b77eb865da633637bd707e741fe59835..ba005cdf72c6bfa8f616806243a43848d80ae91a
--- a/data/tuto1/vtt/tuto1-activite1-vid4-it.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid4-it.vtt
@@ -1,36 +1,65 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.521 --> 00:00:04.312
-Il nostro programma può fare solo
-ciò che è stato addestrato a fare.
+1
+00:00:00.700 --> 00:00:04.423
+IL NOSTRO PROGRAMMA SA FARE
+SOLO CIÒ PER CUI È STATO ADDESTRATO!
 
-00:00:04.708 --> 00:00:07.208
-Il programma può riconoscere solo
-quello che gli abbiamo mostrato.
+2
+00:00:04.693 --> 00:00:07.748
+Il programma può riconoscere
+solo ciò che gli è stato mostrato.
 
-00:00:07.527 --> 00:00:11.390
-Se gli insegniamo a riconoscere i gatti o i cani,
-non sarà in grado di riconoscere i leoni.
+3
+00:00:07.773 --> 00:00:10.159
+Se lo si addestra
+a riconoscere cani e gatti,
 
-00:00:11.458 --> 00:00:14.416
-Ma può dirci
-a quale categoria si avvicina.
+4
+00:00:10.184 --> 00:00:11.846
+non potrà riconoscere dei leoni,
 
-00:00:14.500 --> 00:00:18.042
-I gatti? È intelligente solo
-quanto noi gli abbiamo insegnato ad esserlo.
+5
+00:00:11.871 --> 00:00:14.557
+ma ci potrà dire
+a quale categoria si avvicinano di più,
 
-00:00:18.126 --> 00:00:21.226
-Quando scopriamo qualcosa di nuovo,
-non sappiamo subito cosa sia.
+6
+00:00:14.582 --> 00:00:15.584
+cioè ai gatti.
 
-00:00:21.672 --> 00:00:26.160
-Ma una volta che abbiamo imparato, possiamo riconoscerla
-in diverse posizioni o contesti.
+7
+00:00:15.609 --> 00:00:17.762
+Non sa più di quello
+che gli viene insegnato.
 
-00:00:26.632 --> 00:00:28.250
-È lo stesso per il nostro programma?
+8
+00:00:17.787 --> 00:00:19.402
+Quando scopriamo una cosa nuova,
+
+9
+00:00:19.427 --> 00:00:21.950
+neanche noi sappiamo subito
+di cosa si tratta,
+
+10
+00:00:21.975 --> 00:00:23.451
+ma, una volta appreso cos'è,
+
+11
+00:00:23.476 --> 00:00:26.606
+siamo capaci di riconoscerla
+in posizioni e contesti diversi.
+
+12
+00:00:26.631 --> 00:00:28.544
+Vale lo stesso per il nostro programma?
+
+13
+00:00:28.569 --> 00:00:30.539
+Se gli si insegna a riconoscere i leoni,
+
+14
+00:00:30.564 --> 00:00:32.759
+saprà riconoscere dei leoni di peluche?
 
-00:00:28.792 --> 00:00:32.390
-Se gli insegniamo a riconoscere i leoni,
-riconoscerà i leoni di peluche?
diff --git a/data/tuto1/vtt/tuto1-activite1-vid4-sl.vtt b/data/tuto1/vtt/tuto1-activite1-vid4-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..cafb99f7b5120178efc4ecbde721088949b18ebd
--- /dev/null
+++ b/data/tuto1/vtt/tuto1-activite1-vid4-sl.vtt
@@ -0,0 +1,51 @@
+WEBVTT FILE
+
+WEBVTT
+
+00:00:00.173 --> 00:00:04.423
+Naš program zna narediti samo to,
+kar smo ga naučili!
+
+00:00:04.465 --> 00:00:07.215
+Program prepozna samo tisto,
+kar smo mu pokazali.
+
+00:00:07.256 --> 00:00:09.881
+Če ga naučimo
+prepoznati mačke ali pse,
+
+00:00:09.923 --> 00:00:11.465
+na podobi ne bo znal prepoznati leva,
+
+00:00:11.506 --> 00:00:14.423
+znal pa bo povedati
+v katero kategorijo najverjetneje spada,
+
+00:00:14.465 --> 00:00:15.548
+torej med mačke.
+
+00:00:16.131 --> 00:00:18.006
+Inteligenten je le toliko, kolikor ga naučimo.
+
+00:00:18.048 --> 00:00:19.465
+Ko odkrijemo nekaj novega,
+
+00:00:19.548 --> 00:00:21.423
+tudi ljudje sprva ne vemo, za kaj gre.
+
+00:00:21.465 --> 00:00:22.965
+Ko pa se naučimo,
+
+00:00:23.048 --> 00:00:26.215
+znamo to stvar prepoznati tudi
+v različnih drugih okoliščinah oz. kontekstih.
+
+00:00:26.506 --> 00:00:28.381
+Ali za naš program velja enako?
+
+00:00:28.673 --> 00:00:30.340
+Če ga naučimo prepoznati leva,
+
+00:00:30.423 --> 00:00:32.673
+ali bo znal razlikovati med pravo živaljo in plišasto igračo?
+
diff --git a/data/tuto1/vtt/tuto1-activite1-vid5-en.vtt b/data/tuto1/vtt/tuto1-activite1-vid5-en.vtt
index 2e6e2a9dff58f5cdbdacea019191c10534a12101..cd54f05a4b7b78159a46cfd3dc48f219509fba9b 100644
--- a/data/tuto1/vtt/tuto1-activite1-vid5-en.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid5-en.vtt
@@ -1,41 +1,63 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.167 --> 00:00:01.429
-It can get things wrong!
+1
+00:00:00.048 --> 00:00:01.631
+It may be wrong!
 
-00:00:01.760 --> 00:00:03.178
-It is just a prediction.
+2
+00:00:01.673 --> 00:00:03.340
+It's just a prediction.
 
-00:00:03.992 --> 00:00:06.990
-Being able to recognise things
-in different positions or contexts
+3
+00:00:03.840 --> 00:00:05.906
+Recognising things in different positions
 
-00:00:07.021 --> 00:00:08.541
-is called generalisation.
+4
+00:00:05.931 --> 00:00:08.506
+or contexts is called generalisation,
 
-00:00:08.583 --> 00:00:11.875
-That's what we try to do when we train
-a program to recognise things.
+5
+00:00:08.556 --> 00:00:09.806
+and that's what you try to do
 
-00:00:12.218 --> 00:00:16.040
+6
+00:00:09.840 --> 00:00:12.841
+when you train a programme
+to identify stuff.
+
+7
+00:00:12.872 --> 00:00:16.481
 We are very good at generalising
 because we are quite approximate.
 
-00:00:16.654 --> 00:00:19.104
+8
+00:00:16.506 --> 00:00:19.927
 Our program is very powerful
-because it is very precise.
+because it is very accurate,
 
-00:00:19.490 --> 00:00:21.367
-But it is less good at generalising.
+9
+00:00:19.958 --> 00:00:22.068
+but it is not so good at generalising.
 
-00:00:22.156 --> 00:00:26.660
+10
+00:00:22.161 --> 00:00:24.423
 However, if we train it correctly
-with thousands of varied examples,
 
-00:00:27.047 --> 00:00:29.830
-in some cases, it can be
-more precise than humans.
+11
+00:00:24.465 --> 00:00:26.296
+with thousands of different examples,
+
+12
+00:00:26.321 --> 00:00:29.590
+it can become more precise
+than a human being in certain cases.
+
+13
+00:00:29.840 --> 00:00:33.226
+Today, we have programs
+that can read medical images
+
+14
+00:00:33.251 --> 00:00:35.840
+more reliably than humans.
 
-00:00:29.875 --> 00:00:35.646
-Today, we have programs that can read
-medical images more reliably than humans.
\ No newline at end of file
diff --git a/data/tuto1/vtt/tuto1-activite1-vid5-it.vtt b/data/tuto1/vtt/tuto1-activite1-vid5-it.vtt
old mode 100755
new mode 100644
index aa0c8997bb73329f2ba7a6b0f9caf9cb6f46c9f9..9985dda786b4402ad228e9b051e7b01c4d9c2109
--- a/data/tuto1/vtt/tuto1-activite1-vid5-it.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid5-it.vtt
@@ -1,41 +1,65 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.167 --> 00:00:01.429
-Può sbagliare le cose!
+1
+00:00:00.048 --> 00:00:01.631
+SÌ, PUÒ SBAGLIARSI!
 
-00:00:01.760 --> 00:00:03.178
-È solo una previsione.
+2
+00:00:01.673 --> 00:00:03.340
+Non sono altro che previsioni.
 
-00:00:03.992 --> 00:00:06.990
-Essere in grado di riconoscere le cose
-in diverse posizioni o contesti
+3
+00:00:03.418 --> 00:00:06.893
+L'essere in grado di riconoscere
+le cose in posizioni o contesti diversi
 
-00:00:07.021 --> 00:00:08.541
-si chiama generalizzazione.
+4
+00:00:06.918 --> 00:00:08.523
+si chiama "generalizzazione",
 
-00:00:08.583 --> 00:00:11.875
-Questo è ciò che cerchiamo di fare quando addestriamo
-un programma a riconoscere le cose.
+5
+00:00:08.548 --> 00:00:10.066
+ed è ciò che si cerca di fare
 
-00:00:12.218 --> 00:00:16.040
+6
+00:00:10.091 --> 00:00:12.695
+quando si addestra un programma
+a riconoscere delle cose.
+
+7
+00:00:12.720 --> 00:00:16.326
 Siamo molto bravi a generalizzare
-perché siamo abbastanza approssimativi.
+perché siamo alquanto approssimativi.
 
-00:00:16.654 --> 00:00:19.104
+8
+00:00:16.351 --> 00:00:19.645
 Il nostro programma è molto potente
-perché è molto preciso.
+perché è molto preciso,
+
+9
+00:00:19.670 --> 00:00:21.697
+ma quindi meno bravo a generalizzare.
+
+10
+00:00:21.722 --> 00:00:24.029
+Tuttavia, se lo si addestra correttamente,
+
+11
+00:00:24.054 --> 00:00:26.149
+con migliaia e migliaia di esempi diversi,
 
-00:00:19.490 --> 00:00:21.367
-Ma è meno bravo a generalizzare.
+12
+00:00:26.174 --> 00:00:29.301
+può, in certi casi, essere
+più preciso dell'essere umano.
 
-00:00:22.156 --> 00:00:26.660
-Tuttavia, se lo addestriamo correttamente
-con migliaia di esempi diversi,
+13
+00:00:29.326 --> 00:00:32.541
+Al giorno d'oggi, infatti, esistono
+dei programmi in grado di leggere
 
-00:00:27.047 --> 00:00:29.830
-in alcuni casi, può essere
-più preciso degli umani.
+14
+00:00:32.566 --> 00:00:35.866
+le immagini diagnostiche
+in modo più affidabile degli umani.
 
-00:00:29.875 --> 00:00:35.646
-Oggi, abbiamo programmi che possono leggere
-immagini mediche in modo più affidabile degli umani.
diff --git a/data/tuto1/vtt/tuto1-activite1-vid5-sl.vtt b/data/tuto1/vtt/tuto1-activite1-vid5-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..4a3a9c43850012ed74cc5b8d6a9f36f47689724e
--- /dev/null
+++ b/data/tuto1/vtt/tuto1-activite1-vid5-sl.vtt
@@ -0,0 +1,51 @@
+WEBVTT FILE
+
+WEBVTT
+
+00:00:00.048 --> 00:00:01.631
+Ja, stroj se lahko zmoti!
+
+00:00:01.673 --> 00:00:03.340
+Kar počne, je zgolj predvidevanje.
+
+00:00:03.840 --> 00:00:06.173
+Prepoznavanje stvari v različnih okoliščinah
+
+00:00:06.215 --> 00:00:08.506
+ali kontekstih se imenuje generalizacija.
+
+00:00:08.548 --> 00:00:09.798
+K temu stremite,
+
+00:00:09.840 --> 00:00:11.840
+ko program učite,
+kako naj prepozna določeno stvar.
+
+00:00:11.881 --> 00:00:16.048
+Ljudje smo v generalizaciji dobri,
+ker težimo k približkom.
+
+00:00:16.506 --> 00:00:19.090
+Naš program pa je sicer zelo zmogljiv,
+saj je izjemno natančen,
+
+00:00:19.131 --> 00:00:21.340
+generalizacija pa mu ne gre tako dobro.
+
+00:00:21.881 --> 00:00:24.423
+Vendar, če ga pravilno učimo,
+
+00:00:24.465 --> 00:00:26.840
+tako da mu pokažemo na tisoče različnih primerov,
+
+00:00:26.881 --> 00:00:29.590
+je lahko v nekaterih primerih
+bistveno natančnejši od ljudi.
+
+00:00:29.840 --> 00:00:34.048
+Danes v medicini obstajajo programi,
+
+00:00:34.090 --> 00:00:35.840
+ki so pri interpretiranju izvidov
+uspešnejši kot ljudje.
+
diff --git a/data/tuto1/vtt/tuto1-activite1-vid6-en.vtt b/data/tuto1/vtt/tuto1-activite1-vid6-en.vtt
index 29ffe2ab53c6a0301ca1171f887a289eaa41fc2f..d0aea7cee4d4a14b6c286f8f36682424c740e0a1 100644
--- a/data/tuto1/vtt/tuto1-activite1-vid6-en.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid6-en.vtt
@@ -1,32 +1,43 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.130 --> 00:00:02.054
+1
+00:00:00.046 --> 00:00:02.006
 What can we do with this program?
 
-00:00:02.487 --> 00:00:05.729
-Now we know
+2
+00:00:02.423 --> 00:00:05.798
+Now that we have understood
 how to train artificial intelligence,
 
-00:00:05.791 --> 00:00:08.125
-we can train it with whatever we want.
+3
+00:00:05.840 --> 00:00:08.173
+we can teach it to do whatever we want.
 
-00:00:08.740 --> 00:00:10.778
+4
+00:00:08.590 --> 00:00:10.881
 Can I train a program
 to tell the difference
 
-00:00:10.812 --> 00:00:12.385
+5
+00:00:10.923 --> 00:00:12.715
 between my cup and my glass?
 
-00:00:12.917 --> 00:00:15.099
-Between a closed hand and an open hand?
+6
+00:00:12.763 --> 00:00:15.097
+Or between my closed hand
+and my open hand?
 
-00:00:15.658 --> 00:00:17.792
-Between my blue shirt and a red T-shirt?
+7
+00:00:15.423 --> 00:00:17.756
+My blue shirt and a red t-shirt?
 
-00:00:17.875 --> 00:00:21.242
-You can choose your two categories
+8
+00:00:17.798 --> 00:00:21.465
+You choose your two categories
 and your ten examples for each.
 
-00:00:21.750 --> 00:00:24.195
-Make sure that
-your two categories are distinct.
\ No newline at end of file
+9
+00:00:21.715 --> 00:00:24.090
+Make sure
+your two categories are distinct.
+
diff --git a/data/tuto1/vtt/tuto1-activite1-vid6-it.vtt b/data/tuto1/vtt/tuto1-activite1-vid6-it.vtt
old mode 100755
new mode 100644
index 893c35a3dce937e2f66789ae877c587c79862226..02a92373dee352a0a18f90a283d1b672d01c0f64
--- a/data/tuto1/vtt/tuto1-activite1-vid6-it.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid6-it.vtt
@@ -1,32 +1,40 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.130 --> 00:00:02.054
-Cosa possiamo fare con questo programma?
+1
+00:00:00.048 --> 00:00:02.381
+COSA SI PUÒ FARE CON QUESTO PROGRAMMA?
 
-00:00:02.487 --> 00:00:05.729
-Ora sappiamo
-come addestrare l'intelligenza artificiale,
+2
+00:00:02.423 --> 00:00:05.968
+Ora che abbiamo capito come addestrare
+un'intelligenza artificiale,
 
-00:00:05.791 --> 00:00:08.125
-possiamo addestrarla con qualsiasi cosa vogliamo.
+3
+00:00:05.993 --> 00:00:08.326
+possiamo addestrarla
+con quello che vogliamo.
 
-00:00:08.740 --> 00:00:10.778
+4
+00:00:08.924 --> 00:00:12.530
 Posso addestrare un programma
-per capire la differenza
+a distinguere una tazza da un bicchiere?
 
-00:00:10.812 --> 00:00:12.385
-tra la mia tazza e il mio bicchiere?
+5
+00:00:12.896 --> 00:00:15.230
+O la mia mano chiusa
+dalla mia mano aperta?
 
-00:00:12.917 --> 00:00:15.099
-Tra una mano chiusa e una mano aperta?
+6
+00:00:15.548 --> 00:00:17.937
+O una camicetta blu da una T-shirt rossa?
 
-00:00:15.658 --> 00:00:17.792
-Tra la camicia blu e la maglietta rossa?
+7
+00:00:17.962 --> 00:00:21.306
+Scegli tu due categorie
+e dieci esempi per ciascuna di esse.
 
-00:00:17.875 --> 00:00:21.242
-Potete scegliere le vostre due categorie
-e i vostri dieci esempi per ciascuna.
+8
+00:00:21.331 --> 00:00:24.226
+Assicurati che le due categorie
+siano ben distinte.
 
-00:00:21.750 --> 00:00:24.195
-Assicuratevi che
-le vostre due categorie siano distinte.
\ No newline at end of file
diff --git a/data/tuto1/vtt/tuto1-activite1-vid6-sl.vtt b/data/tuto1/vtt/tuto1-activite1-vid6-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..eb10ab3433f575fdefbc9c86fc71b5b112440bb5
--- /dev/null
+++ b/data/tuto1/vtt/tuto1-activite1-vid6-sl.vtt
@@ -0,0 +1,35 @@
+WEBVTT FILE
+
+WEBVTT
+
+00:00:00.048 --> 00:00:02.006
+Kaj lahko počnemo s takšnim programom?
+
+00:00:02.423 --> 00:00:05.798
+Zdaj ko razumemo,
+kako učiti umetno inteligenco,
+
+00:00:05.840 --> 00:00:08.173
+jo lahko naučimo česarkoli.
+
+00:00:08.590 --> 00:00:10.881
+Lahko program naučim,
+
+00:00:10.923 --> 00:00:12.715
+da prepozna razliko
+med skodelico in kozarcem?
+
+00:00:12.756 --> 00:00:15.090
+Ali med pestjo in odprto dlanjo?
+
+00:00:15.423 --> 00:00:17.756
+Med mojo modro srajco
+in rdečo majico s kratkimi rokavi?
+
+00:00:17.798 --> 00:00:21.465
+Izberite dve kategoriji
+in deset primerov za vsako.
+
+00:00:21.715 --> 00:00:24.090
+Prepričajte se, da sta izbrani kategoriji raznoliki.
+
diff --git a/data/tuto1/vtt/tuto1-activite1-vid7-en.vtt b/data/tuto1/vtt/tuto1-activite1-vid7-en.vtt
index 545d647ced54396c1404e504a5551c9de249f11b..1bad78465d083eadf16b9108c2dd288d6eb2485c 100644
--- a/data/tuto1/vtt/tuto1-activite1-vid7-en.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid7-en.vtt
@@ -1,75 +1,120 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.125 --> 00:00:02.352
-Impressive! But how is it useful?
+1
+00:00:00.131 --> 00:00:02.215
+Impressive, but what is it for?
 
-00:00:02.583 --> 00:00:04.917
-Impressive, right? We agree.
+2
+00:00:02.506 --> 00:00:03.631
+Impressive, right?
 
-00:00:05.001 --> 00:00:07.200
-The program doesn't really see
-our categories.
+3
+00:00:03.673 --> 00:00:04.965
+All right, we agree.
 
-00:00:07.469 --> 00:00:10.521
-But it associates the label we chose
-to the example we showed it.
+4
+00:00:05.006 --> 00:00:07.298
+The program doesn't see our categories,
 
-00:00:10.562 --> 00:00:13.208
-This ability to recognise things
-can be very useful.
+5
+00:00:07.340 --> 00:00:10.506
+but it associates the label we've chosen
+with the examples we show it.
 
-00:00:13.445 --> 00:00:15.910
-We're finding more
-and more AI in our daily lives.
+6
+00:00:10.840 --> 00:00:13.381
+This ability to identify things
+can be very useful,
 
-00:00:15.994 --> 00:00:19.910
-It lets us automatically translate text,
-communicate with voice assistants
-
-00:00:20.210 --> 00:00:22.807
-or even improve medical diagnoses.
-
-00:00:23.083 --> 00:00:26.917
-Computer engineers even use it
-to try and build autonomous cars!
-
-00:00:27.000 --> 00:00:30.146
-But many uses are still to be invented
-to make the world of tomorrow
-
-00:00:30.208 --> 00:00:32.958
-fairer, simpler.
-more beautiful and more sustainable.
-
-00:00:33.042 --> 00:00:36.333
-Helping people
-with disabilities, for example,
+7
+00:00:13.423 --> 00:00:17.011
+and AIs are becoming increasingly common
+in our day-to-day lives.
 
-00:00:36.417 --> 00:00:39.208
-with applications
-that describe the environment
+8
+00:00:17.043 --> 00:00:19.293
+It can automatically translate text,
 
-00:00:39.292 --> 00:00:40.708
-for people with vision impairments,
-
-00:00:40.792 --> 00:00:43.213
-or allowing paralysed people to write,
-
-00:00:43.297 --> 00:00:48.166
-making work simpler and more interesting,
-by automating some repetitive tasks.
-
-00:00:48.250 --> 00:00:53.792
-reducing climate change by making
-more accurate predictions for the future.
-
-00:00:54.109 --> 00:00:58.720
-What about us? Now you know what AI is,
-but what else can we do with it?
+9
+00:00:19.318 --> 00:00:21.173
+communicate with voice assistants
 
-00:00:59.250 --> 00:01:01.739
+10
+00:00:21.215 --> 00:00:22.881
+or improve medical diagnoses.
+
+11
+00:00:22.923 --> 00:00:25.423
+Computer scientists are also using it
+
+12
+00:00:25.465 --> 00:00:26.965
+to try to build autonomous cars,
+
+13
+00:00:27.073 --> 00:00:30.479
+but many uses are still to be invented
+for the world of tomorrow,
+
+14
+00:00:30.565 --> 00:00:32.401
+which will be more beautiful, fairer,
+
+15
+00:00:32.426 --> 00:00:33.943
+more sustainable and simpler.
+
+16
+00:00:33.975 --> 00:00:36.559
+Helping people with reduced autonomy,
+for example,
+
+17
+00:00:36.584 --> 00:00:38.173
+by developing applications
+
+18
+00:00:38.215 --> 00:00:40.631
+that can describe their surroundings
+by seeing us
+
+19
+00:00:40.673 --> 00:00:43.365
+or enabling paralysed people
+to write text.
+
+20
+00:00:43.432 --> 00:00:45.993
+Making work easier and more interesting
+
+21
+00:00:46.018 --> 00:00:48.273
+by automating certain repetitive tasks.
+
+22
+00:00:48.607 --> 00:00:50.076
+Limiting global warming
+
+23
+00:00:50.101 --> 00:00:53.756
+by allowing more accurate predictions
+of the future.
+
+24
+00:00:54.048 --> 00:00:56.256
+Now that we know what AI is,
+
+25
+00:00:56.298 --> 00:00:57.885
+what else can we do with it?
+
+26
+00:00:58.120 --> 00:01:00.620
 What solutions can we invent
-for the world of tomorrow?
+for the future?
+
+27
+00:01:00.979 --> 00:01:04.173
+How can we use it
+in our day-to-day lives to help us?
 
-00:01:01.840 --> 00:01:04.229
-How can we use it every day to help us?
\ No newline at end of file
diff --git a/data/tuto1/vtt/tuto1-activite1-vid7-it.vtt b/data/tuto1/vtt/tuto1-activite1-vid7-it.vtt
old mode 100755
new mode 100644
index 41ba421fbaa10a57cda9a2a50a49eddcfa8f7114..2e8d52e4f49b11ce499cee4789c1ca05f829b83a
--- a/data/tuto1/vtt/tuto1-activite1-vid7-it.vtt
+++ b/data/tuto1/vtt/tuto1-activite1-vid7-it.vtt
@@ -1,75 +1,122 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.125 --> 00:00:02.352
-Impressionante! Ma come è utile?
+1
+00:00:00.131 --> 00:00:02.465
+INCREDIBILE! MA A CHE SERVE?
 
-00:00:02.583 --> 00:00:04.917
-Impressionante, vero? Siamo d'accordo.
+2
+00:00:02.506 --> 00:00:03.573
+Incredibile, no?
 
-00:00:05.001 --> 00:00:07.200
-Il programma non vede davvero
-le nostre categorie.
+3
+00:00:03.598 --> 00:00:04.765
+Beh, siamo d'accordo.
 
-00:00:07.469 --> 00:00:10.521
-Ma associa l'etichetta che abbiamo scelto
-all'esempio che gli abbiamo mostrato.
+4
+00:00:04.790 --> 00:00:07.315
+Il programma non riconosce davvero
+le nostre categorie,
 
-00:00:10.562 --> 00:00:13.208
-Questa capacità di riconoscere le cose
-può essere molto utile.
-
-00:00:13.445 --> 00:00:15.910
-Stiamo trovando sempre più
-IA nella nostra vita quotidiana.
-
-00:00:15.994 --> 00:00:19.910
-Ci permette di tradurre automaticamente il testo,
-comunicare con gli assistenti vocali
-
-00:00:20.210 --> 00:00:22.807
-o persino migliorare le diagnosi mediche.
-
-00:00:23.083 --> 00:00:26.917
-Gli ingegneri informatici lo usano anche
-per provare a costruire auto autonome!
-
-00:00:27.000 --> 00:00:30.146
-Ma molti usi sono ancora da inventare
-per rendere il mondo di domani
-
-00:00:30.208 --> 00:00:32.958
-più giusto, più semplice.
-più bello e più sostenibile.
-
-00:00:33.042 --> 00:00:36.333
-Aiutare le persone
-con disabilità, per esempio,
+5
+00:00:07.340 --> 00:00:10.531
+ma associa l'etichetta che abbiamo scelto
+agli esempi che gli mostriamo.
 
-00:00:36.417 --> 00:00:39.208
-con applicazioni
-che descrivono l'ambiente
-
-00:00:39.292 --> 00:00:40.708
-per persone con problemi di vista,
-
-00:00:40.792 --> 00:00:43.213
-o per permettere alle persone paralizzate di scrivere,
-
-00:00:43.297 --> 00:00:48.166
-rendendo il lavoro più semplice e interessante,
-automatizzando alcuni compiti ripetitivi.
-
-00:00:48.250 --> 00:00:53.792
-ridurre il cambiamento climatico facendo
-previsioni più accurate per il futuro.
-
-00:00:54.109 --> 00:00:58.720
-E noi? Ora sai cos'è l'IA,
-ma cos'altro possiamo fare con essa?
-
-00:00:59.250 --> 00:01:01.739
-Quali soluzioni possiamo inventare
+6
+00:00:10.556 --> 00:00:13.832
+Questa capacità di riconoscere le cose
+può risultare molto utile,
+
+7
+00:00:13.857 --> 00:00:17.206
+infatti troviamo sempre più IA
+nella nostra vita quotidiana.
+
+8
+00:00:17.231 --> 00:00:19.506
+Permette di tradurre
+automaticamente testi,
+
+9
+00:00:19.531 --> 00:00:21.654
+di comunicare con degli assistenti vocali,
+
+10
+00:00:21.679 --> 00:00:23.847
+o addirittura di migliorare
+le diagnosi mediche.
+
+11
+00:00:23.872 --> 00:00:27.948
+Gli informatici la usano anche per provare
+a costruire auto a guida autonoma,
+
+12
+00:00:27.973 --> 00:00:30.369
+ma sono numerosi
+gli utilizzi ancora da inventare
+
+13
+00:00:30.394 --> 00:00:32.708
+per il mondo del domani,
+più bello, più giusto,
+
+14
+00:00:32.733 --> 00:00:34.530
+più sostenibile e più semplice.
+
+15
+00:00:34.555 --> 00:00:36.202
+Aiutare le persone con disabilità,
+
+16
+00:00:36.227 --> 00:00:38.362
+ad esempio, sviluppando applicazioni
+
+17
+00:00:38.387 --> 00:00:41.315
+in grado di descrivere l'ambiente
+che circonda i non vedenti,
+
+18
+00:00:41.340 --> 00:00:44.461
+o di permettere a una persona paralizzata
+di scrivere un testo.
+
+19
+00:00:44.486 --> 00:00:46.767
+Rendere il lavoro
+più semplice e più stimolante,
+
+20
+00:00:46.792 --> 00:00:48.900
+automatizzando i compiti più ripetitivi.
+
+21
+00:00:49.040 --> 00:00:50.822
+Limitare il riscaldamento climatico
+
+22
+00:00:50.847 --> 00:00:54.106
+permettendo, ad esempio,
+di fare previsioni più precise sul futuro.
+
+23
+00:00:54.229 --> 00:00:56.309
+E noi, ora che sappiamo cos'è l'IA,
+
+24
+00:00:56.334 --> 00:00:58.596
+che altro possiamo immaginare
+di fare con essa?
+
+25
+00:00:58.621 --> 00:01:01.406
+Che soluzioni possiamo inventare
 per il mondo di domani?
 
-00:01:01.840 --> 00:01:04.229
-Come possiamo usarlo ogni giorno per aiutarci?
+26
+00:01:01.431 --> 00:01:04.460
+Come possiamo sfruttarla nel quotidiano
+per esserci d'aiuto?
+
diff --git a/data/tuto1/vtt/tuto1-activite1-vid7-sl.vtt b/data/tuto1/vtt/tuto1-activite1-vid7-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..fa557274959d44f2ccee8bbb01fd0a30a29d2b82
--- /dev/null
+++ b/data/tuto1/vtt/tuto1-activite1-vid7-sl.vtt
@@ -0,0 +1,95 @@
+WEBVTT FILE
+
+WEBVTT
+
+00:00:00.131 --> 00:00:02.215
+Izjemno! Ampak čemu služi?
+
+00:00:02.506 --> 00:00:03.631
+Res je izjemno, kajne?
+
+00:00:03.673 --> 00:00:04.965
+Ja, saj se strinjamo.
+
+00:00:05.006 --> 00:00:07.298
+Program v resnici
+ne prepozna naših kategorij,
+
+00:00:07.340 --> 00:00:10.506
+vendar jih poveže z oznakami (opisi),
+s katerimi smo opremili posamezne primere.
+
+00:00:10.840 --> 00:00:13.381
+Sposobnost prepoznavanja stvari
+je lahko zelo uporabna,
+
+00:00:13.423 --> 00:00:15.923
+in umetna inteligenca
+je v naših življenjih vse bolj prisotna.
+
+00:00:15.965 --> 00:00:18.131
+Samodejno lahko prevede besedilo,
+
+00:00:18.173 --> 00:00:21.173
+se sporazumeva z virtualnimi asistenti,
+
+00:00:21.215 --> 00:00:22.881
+ali izboljšuje ustreznost zdravstvenih diagnoz.
+
+00:00:22.923 --> 00:00:25.423
+Računalničarji UI uporabljajo tudi
+
+00:00:25.465 --> 00:00:26.965
+pri zasnovi samovozečih avtomobilov,
+
+00:00:27.006 --> 00:00:29.298
+v prihodnosti pa bomo odkrivali
+še veliko drugih načinov uporabe UI,
+
+00:00:29.340 --> 00:00:31.548
+zaradi katerih bo svet lepši, pravičnejši,
+
+00:00:31.590 --> 00:00:33.006
+bolj trajnosten in preprostejši.
+
+00:00:33.048 --> 00:00:35.173
+Z UI si bodo lahko pomagali invalidi,
+
+00:00:35.215 --> 00:00:38.173
+na primer z aplikacijami
+
+00:00:38.215 --> 00:00:40.631
+za opisovanje okolice
+
+00:00:40.673 --> 00:00:42.923
+ali za pretvorbo govora v besedilo.
+
+00:00:42.965 --> 00:00:45.798
+UI delo nam delo olajša in popestri
+
+00:00:45.840 --> 00:00:48.090
+z avtomatiziranjem
+določenih ponavljajočih se opravil.
+
+00:00:48.298 --> 00:00:51.256
+Z zagotavljanjem
+natančnejših napovedi za prihodnost
+
+00:00:51.298 --> 00:00:53.756
+lahko pomaga tudi pri
+omejevanju vplivov globalnega segrevanja.
+
+00:00:54.048 --> 00:00:56.256
+Kaj še lahko počnemo z umetno inteligenco zdaj,
+
+00:00:56.298 --> 00:00:58.881
+ko vemo, kaj to je?
+
+00:00:58.923 --> 00:01:01.881
+Kakšnih rešitev
+se bomo domislili za prihodnost?
+
+00:01:01.923 --> 00:01:04.173
+Kako si lahko z UI pomagamo
+v vsakdanjem življenju?
+
diff --git a/data/tuto2/vtt/tuto2-activite1-vid1-en.vtt b/data/tuto2/vtt/tuto2-activite1-vid1-en.vtt
index 5f7c5175b30d00eb6d28b369668cd7ccf47fb29b..e7a65f6b93e28684c7106b283ac83ca6a1adc233 100644
--- a/data/tuto2/vtt/tuto2-activite1-vid1-en.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid1-en.vtt
@@ -1,41 +1,52 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.333 --> 00:00:02.031
+1
+00:00:00.090 --> 00:00:02.048
 Algorithms and data.
 
-00:00:02.370 --> 00:00:05.187
-Today, when we talk about AI
-or artificial intelligence,
+2
+00:00:02.256 --> 00:00:05.678
+Today, when we talk
+about AI or artificial intelligence,
 
-00:00:05.229 --> 00:00:07.542
-we're often talking
+3
+00:00:05.703 --> 00:00:08.396
+we are most often talking
 about machine learning.
 
-00:00:07.736 --> 00:00:10.403
-Unlike algorithms,
-which were used previously
+4
+00:00:08.421 --> 00:00:10.546
+Unlike previous algorithms,
 
-00:00:10.729 --> 00:00:14.375
-and which involved describing
-an operation step-by-step,
+5
+00:00:10.571 --> 00:00:14.256
+which were a step-by-step description
+of how to perform an operation,
 
-00:00:14.792 --> 00:00:16.792
-a bit like a recipe,
+6
+00:00:14.367 --> 00:00:15.967
+rather like a recipe,
 
-00:00:16.875 --> 00:00:20.312
-machine learning involves
-training a program
+7
+00:00:16.295 --> 00:00:19.359
+machine learning consists
+of training a programme
 
-00:00:20.363 --> 00:00:22.218
-to make predictions from data.
+8
+00:00:19.384 --> 00:00:21.453
+to make predictions based on data.
 
-00:00:22.654 --> 00:00:26.200
-We use it, for example,
-to predict what a user will like
+9
+00:00:22.536 --> 00:00:25.750
+For example,
+it is used to predict a user's likes
 
-00:00:26.229 --> 00:00:28.562
-based on what they have
-already liked or viewed.
+10
+00:00:25.775 --> 00:00:28.281
+based on
+what they have already liked or watched.
+
+11
+00:00:28.954 --> 00:00:29.937
+Let's test it.
 
-00:00:28.792 --> 00:00:29.875
-Let's test it out!
\ No newline at end of file
diff --git a/data/tuto2/vtt/tuto2-activite1-vid1-it.vtt b/data/tuto2/vtt/tuto2-activite1-vid1-it.vtt
old mode 100755
new mode 100644
index bd21502c5a258c15969b8dc9a08b4a5c43b6cf68..cfbb5f4f5287fa410ed347cd14f0c4523818eaf2
--- a/data/tuto2/vtt/tuto2-activite1-vid1-it.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid1-it.vtt
@@ -1,41 +1,57 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.333 --> 00:00:02.031
-Algoritmi e dati.
+1
+00:00:00.399 --> 00:00:02.100
+ALGORITMI E DATI
 
-00:00:02.370 --> 00:00:05.187
-Oggi, quando parliamo di AI
-o intelligenza artificiale,
+2
+00:00:02.207 --> 00:00:06.100
+Al giorno d'oggi, quando si parla
+d'IA o d'intelligenza artificiale,
 
-00:00:05.229 --> 00:00:07.542
-stiamo spesso parlando
+3
+00:00:06.126 --> 00:00:08.817
+il più delle volte si parla
 di apprendimento automatico.
 
-00:00:07.736 --> 00:00:10.403
-A differenza degli algoritmi,
-che erano usati in precedenza
+4
+00:00:08.842 --> 00:00:11.389
+Al contrario degli algoritmi
+utilizzati finora,
 
-00:00:10.729 --> 00:00:14.375
-e che implicavano la descrizione
-un'operazione passo dopo passo,
+5
+00:00:11.414 --> 00:00:14.637
+che descrivevano passo dopo passo
+come effettuare un'operazione,
 
-00:00:14.792 --> 00:00:16.792
-un po' come una ricetta,
+6
+00:00:14.662 --> 00:00:16.507
+un po' come una ricetta di cucina,
 
-00:00:16.875 --> 00:00:20.312
-l'apprendimento automatico comporta
-l'addestramento di un programma
+7
+00:00:16.532 --> 00:00:19.794
+l'apprendimento automatico
+consiste nell'addestrare un programma
 
-00:00:20.363 --> 00:00:22.218
-per fare previsioni dai dati.
+8
+00:00:19.819 --> 00:00:22.028
+a fare delle previsioni
+sulla base di dati.
 
-00:00:22.654 --> 00:00:26.200
-Lo usiamo, per esempio,
-per predire cosa piacerà ad un utente
+9
+00:00:22.053 --> 00:00:23.441
+Può servire, ad esempio,
 
-00:00:26.229 --> 00:00:28.562
-in base a ciò che gli è
-già piaciuto o visto.
+10
+00:00:23.466 --> 00:00:25.588
+a prevedere cosa può piacere a un utente,
+
+11
+00:00:25.613 --> 00:00:28.939
+in base a cosa gli è piaciuto
+fino ad allora o a cosa ha guardato.
+
+12
+00:00:28.964 --> 00:00:29.973
+Proviamo.
 
-00:00:28.792 --> 00:00:29.875
-Mettiamolo alla prova!
\ No newline at end of file
diff --git a/data/tuto2/vtt/tuto2-activite1-vid1-sl.vtt b/data/tuto2/vtt/tuto2-activite1-vid1-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..4b622f68556b744fd3c187f0a60e00ef315cd0d8
--- /dev/null
+++ b/data/tuto2/vtt/tuto2-activite1-vid1-sl.vtt
@@ -0,0 +1,42 @@
+WEBVTT FILE
+
+WEBVTT
+
+00:00:00.090 --> 00:00:02.048
+Algoritmi in podatki.
+
+00:00:02.256 --> 00:00:05.298
+Ko danes govorimo
+o umetni inteligenci,
+
+00:00:05.340 --> 00:00:07.506
+imamo večinoma v mislih strojno učenje.
+
+00:00:07.548 --> 00:00:10.340
+V nasprotju z dosedanjimi algoritmi,
+
+00:00:10.715 --> 00:00:14.215
+ki kakor kuharski recept
+s točno določenim zaporedjem korakov opisujejo,
+
+00:00:14.756 --> 00:00:16.840
+kako izvesti določeno opravilo,
+
+00:00:16.881 --> 00:00:20.131
+je bistvo strojnega učenja v tem,
+da program naučimo predvidevanja
+
+00:00:20.173 --> 00:00:22.131
+na podlagi podatkov.
+
+00:00:22.631 --> 00:00:26.340
+Na primer, program predvidi, kaj bo uporabniku všeč
+
+00:00:26.381 --> 00:00:28.756
+na podlagi tega,
+kar mu je bilo všeč v preteklosti
+ali na podlagi že ogledanih vsebin.
+
+00:00:28.798 --> 00:00:29.756
+Preizkusimo.
+
diff --git a/data/tuto2/vtt/tuto2-activite1-vid2-en.vtt b/data/tuto2/vtt/tuto2-activite1-vid2-en.vtt
index ae6399e1abb9deaa5ade8d2d7761cb97a56182ae..276a15e98b224791daa5d424cabe2046343a540e 100644
--- a/data/tuto2/vtt/tuto2-activite1-vid2-en.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid2-en.vtt
@@ -1,44 +1,70 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.167 --> 00:00:02.757
-Sexist? What if it is our data?
+1
+00:00:00.090 --> 00:00:01.131
+Sexist?
 
-00:00:03.250 --> 00:00:06.250
-You surely noticed
+2
+00:00:01.156 --> 00:00:02.718
+What if it was our data...
+
+3
+00:00:02.743 --> 00:00:05.062
+{\an5}You may have noticed
 that our model recognised
 
-00:00:06.291 --> 00:00:10.390
-everyone with long hair as women
-and everyone with short hair as men.
+4
+00:00:05.087 --> 00:00:07.109
+all people with long hair as women
+
+5
+00:00:07.178 --> 00:00:09.101
+and all those with short hair as men.
+
+6
+00:00:09.126 --> 00:00:10.335
+Without realising it,
+
+7
+00:00:10.360 --> 00:00:12.407
+we provided it with data that was sorted
 
-00:00:10.917 --> 00:00:15.292
-Without realising, we provided data
-that was sorted by criteria
+8
+00:00:12.432 --> 00:00:15.743
+{\an5}according to criteria
+other than those we had defined.
 
-00:00:15.376 --> 00:00:16.751
-other than what we had defined.
+9
+00:00:15.768 --> 00:00:19.666
+{\an5}How we sorted the data was influenced
+by our perception of the problem.
 
-00:00:17.432 --> 00:00:20.410
-The data sorting was influenced
-by our perception of the issue.
+10
+00:00:19.691 --> 00:00:22.557
+{\an5}When this happens,
+we say the data is biased.
 
-00:00:20.836 --> 00:00:23.660
-When that happens,
-we say that the data is biased.
+11
+00:00:22.582 --> 00:00:25.361
+{\an5}AI algorithms have been around
+for a long time,
 
-00:00:23.917 --> 00:00:26.080
-AI algorithms have existed for a long time
+12
+00:00:25.386 --> 00:00:28.769
+{\an5}but they didn't work so well
+because we lacked labelled data.
 
-00:00:26.322 --> 00:00:29.267
-but they didn't work as well
-because we don't have labelled data.
+13
+00:00:28.794 --> 00:00:31.527
+{\an5}Today,
+data is available in large quantities.
 
-00:00:29.739 --> 00:00:31.994
-Today, a lot of data is available.
+14
+00:00:31.552 --> 00:00:33.796
+That's why AI is so widespread.
 
-00:00:32.429 --> 00:00:34.410
-And that's why AI is so widespread.
+15
+00:00:33.821 --> 00:00:38.250
+{\an5}However, we have to be careful
+because this data can contain biases.
 
-00:00:34.750 --> 00:00:37.817
-But you should be careful
-because the data may be biased.
\ No newline at end of file
diff --git a/data/tuto2/vtt/tuto2-activite1-vid2-it.vtt b/data/tuto2/vtt/tuto2-activite1-vid2-it.vtt
old mode 100755
new mode 100644
index dd077b41232030f2604f9b8bef1b083c1b77b9ac..7c30f4affd727bedba5769122cbfc1d0825361d9
--- a/data/tuto2/vtt/tuto2-activite1-vid2-it.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid2-it.vtt
@@ -1,44 +1,69 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.167 --> 00:00:02.757
-Sessista? E se fossero i nostri dati?
+1
+00:00:00.090 --> 00:00:02.965
+SESSISTA? E SE FOSSERO I NOSTRI DATI...
 
-00:00:03.250 --> 00:00:06.250
-Avrete sicuramente notato
-che il nostro modello ha riconosciuto
+2
+00:00:03.006 --> 00:00:05.413
+Sicuramente avrai potuto notare
+che il nostro modello
 
-00:00:06.291 --> 00:00:10.390
-tutti quelli con i capelli lunghi come donne
-e tutti quelli con i capelli corti come uomini.
+3
+00:00:05.438 --> 00:00:08.440
+riconosceva tutte le persone
+con i capelli lunghi come donne
 
-00:00:10.917 --> 00:00:15.292
-Senza rendercene conto, abbiamo fornito dati
-che erano ordinati per criteri
+4
+00:00:08.465 --> 00:00:10.748
+e tutte quelle con i capelli corti
+come uomini.
 
-00:00:15.376 --> 00:00:16.751
-diversi da quelli che avevamo definito.
+5
+00:00:10.773 --> 00:00:12.002
+Senza rendercene conto,
 
-00:00:17.432 --> 00:00:20.410
+6
+00:00:12.027 --> 00:00:13.182
+gli abbiamo fornito dei dati
+
+7
+00:00:13.207 --> 00:00:16.982
+organizzati secondo criteri diversi
+da quelli che noi avevamo definito.
+
+8
+00:00:17.007 --> 00:00:20.936
 L'ordinamento dei dati è stato influenzato
 dalla nostra percezione del problema.
 
-00:00:20.836 --> 00:00:23.660
-Quando questo accade,
-diciamo che i dati sono distorti.
+9
+00:00:20.961 --> 00:00:23.561
+Quando succede,
+si dice che i dati sono distorti.
+
+10
+00:00:23.586 --> 00:00:26.241
+Gli algoritmi d'IA
+esistono da molto tempo,
 
-00:00:23.917 --> 00:00:26.080
-Gli algoritmi di IA esistono da molto tempo
+11
+00:00:26.266 --> 00:00:29.525
+ma non funzionavano così bene
+perché non avevamo dati etichettati.
 
-00:00:26.322 --> 00:00:29.267
-ma non hanno funzionato così bene
-perché non abbiamo dati etichettati.
+12
+00:00:29.550 --> 00:00:32.235
+Oggi, i dati sono disponibili
+e in grande quantità.
 
-00:00:29.739 --> 00:00:31.994
-Oggi sono disponibili molti dati.
+13
+00:00:32.272 --> 00:00:34.727
+È per questo motivo
+che l'IA è così diffusa.
 
-00:00:32.429 --> 00:00:34.410
-Ed è per questo che l'IA è così diffusa.
+14
+00:00:34.752 --> 00:00:37.919
+Ma bisogna fare attenzione,
+perché i dati possono essere distorti.
 
-00:00:34.750 --> 00:00:37.817
-Ma bisogna fare attenzione
-perché i dati potrebbero essere distorti.
\ No newline at end of file
diff --git a/data/tuto2/vtt/tuto2-activite1-vid2-sl.vtt b/data/tuto2/vtt/tuto2-activite1-vid2-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..353d414943bccd9201544ec64c1bd7c623f9319a
--- /dev/null
+++ b/data/tuto2/vtt/tuto2-activite1-vid2-sl.vtt
@@ -0,0 +1,56 @@
+WEBVTT FILE
+
+WEBVTT
+
+00:00:00.090 --> 00:00:01.131
+Seksistično?
+
+00:00:01.423 --> 00:00:02.673
+In če so takšni tudi naši podatki?
+
+00:00:03.298 --> 00:00:06.215
+Najbrž ste opazili,
+
+00:00:06.256 --> 00:00:08.465
+da je naš model vse osebe z dolgimi lasmi
+označil kot ženske,
+
+00:00:08.506 --> 00:00:10.506
+tiste s kratkimi pa kot moške.
+
+00:00:10.881 --> 00:00:12.090
+Ne da bi se zavedali,
+
+00:00:12.465 --> 00:00:14.548
+smo modelu posredovali podatke,
+
+00:00:14.590 --> 00:00:16.673
+razvrščene na podlagi drugačnega
+kriterija, kot smo ga izbrali.
+
+00:00:17.381 --> 00:00:20.340
+Na razvrščanje podatkov
+je vplivalo naše dojemanje problema (okoliščin).
+
+00:00:20.798 --> 00:00:23.506
+Ko se zgodi kaj takega,
+rečemo, da so podatki pristranski.
+
+00:00:23.923 --> 00:00:26.131
+Algoritmi UI obstajajo že dolgo,
+
+00:00:26.340 --> 00:00:29.340
+vendar včasih niso bili tako učinkoviti,
+ker podatki niso bili označeni.
+
+00:00:29.798 --> 00:00:31.965
+Danes so nam na voljo
+velike količine podatkov.
+
+00:00:32.423 --> 00:00:34.340
+Prav zato je UI tako razširjena.
+
+00:00:34.756 --> 00:00:37.923
+Vseeno moramo biti previdni,
+saj lahko podatki vsebujejo predsodke (so pristranski).
+
diff --git a/data/tuto2/vtt/tuto2-activite1-vid3-en.vtt b/data/tuto2/vtt/tuto2-activite1-vid3-en.vtt
index 7b38c3d6967c734324d42f84bfbdb2f1cbb405ff..a6498a5ec37d05fb2f78e9b14f1d45df3c0ff9eb 100644
--- a/data/tuto2/vtt/tuto2-activite1-vid3-en.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid3-en.vtt
@@ -1,29 +1,43 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.292 --> 00:00:02.386
+1
+00:00:00.423 --> 00:00:02.375
 The art of data preparation.
 
-00:00:02.789 --> 00:00:07.083
-Preparing the data for AI
+2
+00:00:02.804 --> 00:00:06.937
+Preparing data to create AI
 is at least 70% of the work.
 
-00:00:07.167 --> 00:00:09.958
+3
+00:00:07.242 --> 00:00:09.953
 Data is very important
-so a program can learn.
+in machine learning.
 
-00:00:10.392 --> 00:00:13.870
+4
+00:00:10.215 --> 00:00:11.490
 As we saw earlier,
-the machine only learns what we show it.
 
-00:00:14.208 --> 00:00:18.470
-We therefore have to spend time selecting
-the data and preparing it properly
+5
+00:00:11.515 --> 00:00:13.983
+the machine only learns
+from what we show it.
 
-00:00:18.521 --> 00:00:19.875
+6
+00:00:14.008 --> 00:00:17.271
+So you have to take the time
+to carefully select this data
+
+7
+00:00:17.296 --> 00:00:19.991
+and prepare it correctly
 to get good results.
 
-00:00:20.321 --> 00:00:22.658
-But it is not always as easy as it looks!
+8
+00:00:20.298 --> 00:00:22.840
+This is not always as easy as it sounds.
+
+9
+00:00:23.129 --> 00:00:25.587
+Let's take a closer look and experiment.
 
-00:00:23.246 --> 00:00:25.796
-Let's take a closer look and experiment!
\ No newline at end of file
diff --git a/data/tuto2/vtt/tuto2-activite1-vid3-it.vtt b/data/tuto2/vtt/tuto2-activite1-vid3-it.vtt
old mode 100755
new mode 100644
index dec0a32bdc38265c832d9b5c76cbd9337139a9e3..31dc151ddf285dd1b56479829dd9dfb23e4a5eaf
--- a/data/tuto2/vtt/tuto2-activite1-vid3-it.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid3-it.vtt
@@ -1,29 +1,47 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.292 --> 00:00:02.386
-L'arte della preparazione dei dati.
+1
+00:00:00.340 --> 00:00:02.266
+L'ARTE DI PREPARARE I DATI
 
-00:00:02.789 --> 00:00:07.083
-Preparare i dati per l'IA
-è almeno il 70% del lavoro.
+2
+00:00:02.606 --> 00:00:04.521
+Preparare i dati per creare un'IA
 
-00:00:07.167 --> 00:00:09.958
+3
+00:00:04.546 --> 00:00:07.140
+corrisponde almeno al 70% del lavoro.
+
+4
+00:00:07.165 --> 00:00:10.188
 I dati sono molto importanti
-perché un programma possa imparare.
+per l'apprendimento di un programma.
+
+5
+00:00:10.239 --> 00:00:11.970
+Come abbiamo visto precedentemente,
 
-00:00:10.392 --> 00:00:13.870
-Come abbiamo visto prima,
-la macchina impara solo ciò che le mostriamo.
+6
+00:00:11.995 --> 00:00:14.363
+la macchina impara solo
+da quello che le mostriamo.
 
-00:00:14.208 --> 00:00:18.470
-Dobbiamo quindi dedicare del tempo a selezionare
-i dati e prepararli adeguatamente
+7
+00:00:14.388 --> 00:00:17.943
+Bisogna quindi prendersi il tempo
+di selezionare bene questi dati
 
-00:00:18.521 --> 00:00:19.875
+8
+00:00:17.968 --> 00:00:21.121
+e di prepararli correttamente
 per ottenere buoni risultati.
 
-00:00:20.321 --> 00:00:22.658
-Ma non è sempre così facile come sembra!
+9
+00:00:21.146 --> 00:00:23.254
+Ma non è sempre così facile come sembra.
+
+10
+00:00:23.279 --> 00:00:26.366
+Diamo uno sguardo più da vicino
+e facciamo un esperimento.
 
-00:00:23.246 --> 00:00:25.796
-Diamo un'occhiata più da vicino e sperimentiamo!
\ No newline at end of file
diff --git a/data/tuto2/vtt/tuto2-activite1-vid3-sl.vtt b/data/tuto2/vtt/tuto2-activite1-vid3-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..d28f676008cb8cf495149ccd1db796ebb84412dd
--- /dev/null
+++ b/data/tuto2/vtt/tuto2-activite1-vid3-sl.vtt
@@ -0,0 +1,36 @@
+WEBVTT FILE
+
+WEBVTT
+
+00:00:00.423 --> 00:00:02.090
+Umetnost priprave podatkov.
+
+00:00:02.548 --> 00:00:07.131
+Pri ustvarjanju UI
+70% dela predstavlja priprava podatkov.
+
+00:00:07.173 --> 00:00:10.173
+Podatki
+so pri strojnem učenju zelo pomembni.
+
+00:00:10.215 --> 00:00:11.756
+Kot smo videli prej,
+
+00:00:11.798 --> 00:00:13.798
+se stroj uči samo na podlagi tega,
+kar mu pokažemo,
+
+00:00:14.215 --> 00:00:16.965
+zato si je treba vzeti čas
+in skrbno izbrati podatke
+
+00:00:17.006 --> 00:00:19.798
+ter jih ustrezno pripraviti,
+saj bomo le tako dobili dobre rezultate.
+
+00:00:20.298 --> 00:00:22.840
+To pa ni vedno tako preprosto, kot se sliši.
+
+00:00:23.215 --> 00:00:25.673
+Poglejmo podrobno in preizkusimo.
+
diff --git a/data/tuto2/vtt/tuto2-activite1-vid4-en.vtt b/data/tuto2/vtt/tuto2-activite1-vid4-en.vtt
index 6826e9d35652612aa3e4f9ea47ec99704f2a4131..2b6ca5e43378b40ef569f30db494d14d99beec76 100644
--- a/data/tuto2/vtt/tuto2-activite1-vid4-en.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid4-en.vtt
@@ -1,42 +1,65 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.542 --> 00:00:02.023
-Biases in data.
+1
+00:00:00.465 --> 00:00:02.164
+Data bias.
 
-00:00:02.574 --> 00:00:04.533
-The program doesn't see things how we do.
+2
+00:00:02.492 --> 00:00:04.606
+The program doesn't see like we do.
 
-00:00:04.617 --> 00:00:07.080
-It doesn't know
-the concept of man and woman.
+3
+00:00:04.631 --> 00:00:07.092
+It does not know what a man or a woman is.
 
-00:00:07.164 --> 00:00:10.250
+4
+00:00:07.117 --> 00:00:10.640
 It makes what are called
 statistical approximations.
 
-00:00:10.984 --> 00:00:15.124
-Is this image closer
-to images labelled "women"
+5
+00:00:10.879 --> 00:00:12.933
+Is this image statistically closer
 
-00:00:15.208 --> 00:00:16.833
-or images labelled "men"?
+6
+00:00:12.958 --> 00:00:14.754
+to images labelled "female"
 
-00:00:16.917 --> 00:00:18.204
-What does it see?
+7
+00:00:14.868 --> 00:00:16.590
+or to those labelled "male"?
 
-00:00:18.458 --> 00:00:20.852
-The blue background, skin colour?
+8
+00:00:16.798 --> 00:00:17.840
+What did it see?
 
-00:00:21.208 --> 00:00:22.542
+9
+00:00:18.081 --> 00:00:19.238
+The blue background?
+
+10
+00:00:19.673 --> 00:00:20.923
+The skin colour?
+
+11
+00:00:21.173 --> 00:00:22.659
 A pair of glasses?
 
-00:00:22.836 --> 00:00:26.305
-The way we choose our input data is key.
+12
+00:00:23.073 --> 00:00:26.057
+How we choose our input data
+is fundamental.
+
+13
+00:00:26.368 --> 00:00:29.368
+Choosing your data
+is a big responsibility.
+
+14
+00:00:30.423 --> 00:00:32.976
+Let's now try to correct our data set
 
-00:00:27.042 --> 00:00:29.917
-Choosing the data is known
-to be a big responsibility.
+15
+00:00:33.001 --> 00:00:34.668
+to eliminate any bias.
 
-00:00:30.495 --> 00:00:34.871
-Now, let's try to correct
-our data set to eliminate the biases.
\ No newline at end of file
diff --git a/data/tuto2/vtt/tuto2-activite1-vid4-it.vtt b/data/tuto2/vtt/tuto2-activite1-vid4-it.vtt
old mode 100755
new mode 100644
index 15183557b12988ec95cf547f4cbf35d7b17034b8..7ede53d14f60a3d39c5b1207cddbcbc04fa8c472
--- a/data/tuto2/vtt/tuto2-activite1-vid4-it.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid4-it.vtt
@@ -1,42 +1,62 @@
-WEBVTT
-
-00:00:00.542 --> 00:00:02.023
-Bias nei dati.
-
-00:00:02.574 --> 00:00:04.533
-Il programma non vede le cose come le vediamo noi.
-
-00:00:04.617 --> 00:00:07.080
-Non conosce
-il concetto di uomo e donna.
-
-00:00:07.164 --> 00:00:10.250
-Fa quelle che sono chiamate
-approssimazioni statistiche.
-
-00:00:10.984 --> 00:00:15.124
-Questa immagine è più vicina
-alle immagini etichettate come "donne"
-
-00:00:15.208 --> 00:00:16.833
-o alle immagini etichettate "uomini"?
-
-00:00:16.917 --> 00:00:18.204
-Cosa vede?
+WEBVTT FILE
+
+1
+00:00:00.646 --> 00:00:02.293
+PREGIUDIZI NEI DATI
+
+2
+00:00:02.514 --> 00:00:04.360
+Il programma non vede come noi.
+
+3
+00:00:04.385 --> 00:00:06.986
+Non conosce il concetto
+di "uomo" o di "donna".
+
+4
+00:00:07.131 --> 00:00:10.351
+Fa quelle che chiamiamo
+"approssimazioni statistiche".
+
+5
+00:00:10.838 --> 00:00:13.832
+Questa immagine si avvicina
+statisticamente di più
+
+6
+00:00:13.857 --> 00:00:18.088
+alle immagini etichettate come "donna"
+o alle immagini etichettate come "uomo"?
+
+7
+00:00:18.113 --> 00:00:19.240
+Che cosa ha visto?
+
+8
+00:00:19.265 --> 00:00:21.434
+Lo sfondo blu?
+Il colore della pelle?
+
+9
+00:00:21.459 --> 00:00:22.667
+Un paio di occhiali?
 
-00:00:18.458 --> 00:00:20.852
-Lo sfondo blu, il colore della pelle?
+10
+00:00:22.780 --> 00:00:26.420
+Il modo in cui scegliamo
+i nostri dati di input è fondamentale.
 
-00:00:21.208 --> 00:00:22.542
-Un paio di occhiali?
+11
+00:00:26.718 --> 00:00:30.013
+Scegliere i dati si sta rivelando
+una grande responsabilità.
 
-00:00:22.836 --> 00:00:26.305
-Il modo in cui scegliamo i nostri dati di input è fondamentale.
+12
+00:00:30.423 --> 00:00:32.999
+Proviamo ora a correggere
+il nostro set di dati
 
-00:00:27.042 --> 00:00:29.917
-La scelta dei dati è nota
-essere una grande responsabilità.
+13
+00:00:33.024 --> 00:00:34.774
+per eliminare tutti i pregiudizi.
 
-00:00:30.495 --> 00:00:34.871
-Ora, proviamo a correggere
-il nostro set di dati per eliminare le distorsioni.
diff --git a/data/tuto2/vtt/tuto2-activite1-vid4-sl.vtt b/data/tuto2/vtt/tuto2-activite1-vid4-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..863c007669ecbe9d378756c936f6090f763562df
--- /dev/null
+++ b/data/tuto2/vtt/tuto2-activite1-vid4-sl.vtt
@@ -0,0 +1,53 @@
+WEBVTT FILE
+
+WEBVTT
+
+00:00:00.465 --> 00:00:02.173
+Pristranski podatki.
+
+00:00:02.215 --> 00:00:04.215
+Program ne "vidi" na enak način kot ljudje.
+
+00:00:04.631 --> 00:00:06.673
+Ne ve, kaj je to 'moški' ali 'ženska'.
+
+00:00:07.048 --> 00:00:10.090
+Namesto tega uporablja
+tako imenovane statistične približke.
+
+00:00:11.090 --> 00:00:13.506
+Je ta podoba
+statistično bolj podobna tistim,
+
+00:00:13.548 --> 00:00:15.256
+ki so označene z opisom "ženska",
+
+00:00:15.298 --> 00:00:16.756
+ali tistim, ki so označene z opisom "moški"?
+
+00:00:16.798 --> 00:00:17.840
+Kaj je videl oz. zaznal program?
+
+00:00:18.340 --> 00:00:19.340
+Modro ozadje?
+
+00:00:19.673 --> 00:00:20.923
+Barvo kože?
+
+00:00:21.173 --> 00:00:22.506
+Očala?
+
+00:00:22.548 --> 00:00:26.173
+Način, kako izberemo vhodne podatke,
+je bistvenega pomena.
+
+00:00:26.923 --> 00:00:29.923
+Ko izbiramo podatke,
+je na nas velika in pomembna odgovornost.
+
+00:00:30.423 --> 00:00:33.131
+Poskušajmo sedaj popraviti naš nabor podatkov tako,
+
+00:00:33.173 --> 00:00:34.840
+da ne bo več pristranski.
+
diff --git a/data/tuto2/vtt/tuto2-activite1-vid5-en.vtt b/data/tuto2/vtt/tuto2-activite1-vid5-en.vtt
index cdd16c659cecfd758555efb4d89234f4b3286dfb..654b6731552acf1383761064fa3633b6c4fd7700 100644
--- a/data/tuto2/vtt/tuto2-activite1-vid5-en.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid5-en.vtt
@@ -1,38 +1,60 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.125 --> 00:00:02.500
+1
+00:00:00.090 --> 00:00:02.215
 Mastering data sets.
 
-00:00:02.738 --> 00:00:06.062
+2
+00:00:02.404 --> 00:00:05.984
 Artificial intelligence can only recognise
-what we've taught it.
+what it has been taught.
 
-00:00:06.375 --> 00:00:10.375
-The data used to train the program has
-a strong influence on the results.
+3
+00:00:06.118 --> 00:00:07.968
+The data you train the programme with
 
-00:00:10.458 --> 00:00:14.328
+4
+00:00:07.993 --> 00:00:10.553
+therefore has a strong influence
+on the results.
+
+5
+00:00:10.578 --> 00:00:14.357
 Mastering the data is a key element
-to master this technology
+in mastering this technology
 
-00:00:14.412 --> 00:00:16.104
+6
+00:00:14.382 --> 00:00:15.965
 and the results you get from it.
 
-00:00:16.150 --> 00:00:19.851
-You should always be vigilant when looking
-at the results from an AI program
+7
+00:00:16.105 --> 00:00:17.590
+We must always be careful
+
+8
+00:00:17.615 --> 00:00:19.965
+when looking at an AI program's results
+
+9
+00:00:20.082 --> 00:00:23.832
+and ask ourselves
+what data was used to train it.
+
+10
+00:00:24.006 --> 00:00:25.148
+Deliberately or not,
 
-00:00:20.250 --> 00:00:23.916
-and always ask
-where the data used to train it is from.
+11
+00:00:25.173 --> 00:00:28.131
+it may contain
+mechanically reproduced biases
 
-00:00:24.000 --> 00:00:28.521
-Deliberately or not, it can contain biases
-that are mechanically reproduced
+12
+00:00:28.204 --> 00:00:30.231
+that have significant consequences,
 
-00:00:28.574 --> 00:00:30.291
-and which have big consequences.
+13
+00:00:30.256 --> 00:00:33.465
+for recruitment, for example,
+or access to credit.
 
-00:00:30.375 --> 00:00:33.542
-For job recruitment for example,
-or even access to a loan.
\ No newline at end of file
diff --git a/data/tuto2/vtt/tuto2-activite1-vid5-it.vtt b/data/tuto2/vtt/tuto2-activite1-vid5-it.vtt
old mode 100755
new mode 100644
index cf27bb8cccf5e2aafb500755df6128b4d622fcbd..bd5b9fa285eaaa0ce3d60cf9ee93c15fe23c9de8
--- a/data/tuto2/vtt/tuto2-activite1-vid5-it.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid5-it.vtt
@@ -1,38 +1,58 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.125 --> 00:00:02.500
-Padroneggiare le serie di dati.
+1
+00:00:00.090 --> 00:00:02.465
+PADRONEGGIARE I SET DI DATI
 
-00:00:02.738 --> 00:00:06.062
-L'intelligenza artificiale può riconoscere solo
-ciò che le abbiamo insegnato.
+2
+00:00:02.506 --> 00:00:06.153
+L'intelligenza artificiale riconosce
+solo ciò che le viene insegnato.
 
-00:00:06.375 --> 00:00:10.375
-I dati usati per addestrare il programma hanno
-una forte influenza sui risultati.
+3
+00:00:06.178 --> 00:00:08.642
+I dati con i quali
+viene addestrato un programma
 
-00:00:10.458 --> 00:00:14.328
-Padroneggiare i dati è un elemento chiave
+4
+00:00:08.667 --> 00:00:10.928
+influiscono fortemente sui risultati.
+
+5
+00:00:10.953 --> 00:00:15.178
+Padroneggiare i dati è un fattore chiave
 per padroneggiare questa tecnologia
 
-00:00:14.412 --> 00:00:16.104
-e i risultati che si ottengono da essa.
+6
+00:00:15.203 --> 00:00:16.725
+e i risultati che se ne ricavano.
+
+7
+00:00:16.750 --> 00:00:18.193
+Bisogna sempre fare attenzione
+
+8
+00:00:18.218 --> 00:00:20.933
+quando si osservano i risultati
+di un programma d'IA
 
-00:00:16.150 --> 00:00:19.851
-Si dovrebbe sempre essere vigili quando si guarda
-i risultati di un programma AI
+9
+00:00:20.958 --> 00:00:24.436
+e chiedersi sempre a partire
+da quali dati è stato addestrato.
 
-00:00:20.250 --> 00:00:23.916
-e chiedere sempre
-da dove provengono i dati usati per addestrarlo.
+10
+00:00:24.461 --> 00:00:27.066
+Volontariamente o no,
+può contenere dei pregiudizi
 
-00:00:24.000 --> 00:00:28.521
-Deliberatamente o no, può contenere distorsioni
-che sono riprodotti meccanicamente
+11
+00:00:27.091 --> 00:00:30.764
+che vengono riprodotti meccanicamente
+e che hanno conseguenze importanti,
 
-00:00:28.574 --> 00:00:30.291
-e che hanno grandi conseguenze.
+12
+00:00:30.789 --> 00:00:33.593
+per un'assunzione, ad esempio,
+o per l'accesso al credito.
 
-00:00:30.375 --> 00:00:33.542
-Per il reclutamento del lavoro, per esempio,
-o anche per l'accesso a un prestito.
diff --git a/data/tuto2/vtt/tuto2-activite1-vid5-sl.vtt b/data/tuto2/vtt/tuto2-activite1-vid5-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..d9c187fa156dc28429871794c397ac10a48c0f80
--- /dev/null
+++ b/data/tuto2/vtt/tuto2-activite1-vid5-sl.vtt
@@ -0,0 +1,48 @@
+WEBVTT FILE
+
+WEBVTT
+
+00:00:00.090 --> 00:00:02.215
+Obvladovanje zbirk podatkov.
+
+00:00:02.506 --> 00:00:06.256
+Umetna inteligenca prepozna samo tisto,
+kar jo naučimo prepoznati.
+
+00:00:06.298 --> 00:00:08.423
+Podatki, ki jih uporabljamo za učenje programa,
+
+00:00:08.465 --> 00:00:10.340
+bistveno vplivajo na rezultate.
+
+00:00:10.381 --> 00:00:14.423
+Nujno je, da dobro poznamo delo s podatki,
+saj lahko le tako obvladamo te tehnologije
+
+00:00:14.465 --> 00:00:16.048
+in žanjemo rezultate njihovega delovanja.
+
+00:00:16.090 --> 00:00:17.506
+Pri interpretaciji rezultatov,
+nastalih s pomočjo programov UI
+
+00:00:17.548 --> 00:00:19.965
+moramo biti vedno previdni.
+
+00:00:20.215 --> 00:00:23.965
+Vprašati se moramo, kateri podatki
+so bili uporabljeni v fazi učenja programa.
+
+00:00:24.006 --> 00:00:25.256
+Načrtno ali ne, program lahko vsebuje
+
+00:00:25.298 --> 00:00:28.256
+mehansko reproducirane pristranske podatke,
+
+00:00:28.298 --> 00:00:30.215
+ki pomembno vplivajo na rezultate,
+na primer, v postopkih iskanja novih kadrov
+
+00:00:30.256 --> 00:00:33.465
+ali pri obravnavanju kreditnih vlog.
+
diff --git a/data/tuto2/vtt/tuto2-activite1-vid6-en.vtt b/data/tuto2/vtt/tuto2-activite1-vid6-en.vtt
index 7b6546ef1e239c6b004d545e15ef1c00f4b7b41d..b9aeae12f4511ddd3023c4265e6d98e2ff84b381 100644
--- a/data/tuto2/vtt/tuto2-activite1-vid6-en.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid6-en.vtt
@@ -1,64 +1,101 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.458 --> 00:00:03.659
-What if we played with the data
+1
+00:00:00.718 --> 00:00:03.673
+What if we play with the data
 to trick the machine?
 
-00:00:04.296 --> 00:00:08.405
+2
+00:00:04.232 --> 00:00:07.023
 The program will make predictions
-in the categories that we defined
-
-00:00:08.437 --> 00:00:10.160
-and the examples we showed it.
-
-00:00:10.458 --> 00:00:12.750
-Once it is trained, we can trick it.
-
-00:00:12.821 --> 00:00:17.578
-For example, certain jewellery
-can deceive facial recognition
-
-00:00:18.160 --> 00:00:20.406
-or even some signs with patterns
-
-00:00:21.105 --> 00:00:24.398
-can get around
-automatic video surveillance.
-
-00:00:24.750 --> 00:00:27.916
-But as we know, we can also
-trick the machine with our data.
-
-00:00:28.000 --> 00:00:32.406
+based on the categories
+
+3
+00:00:07.048 --> 00:00:09.953
+we have defined and the examples
+we have shown to it.
+
+4
+00:00:10.406 --> 00:00:13.169
+Once trained, it is possible to trick it.
+
+5
+00:00:13.380 --> 00:00:14.333
+For example,
+
+6
+00:00:14.358 --> 00:00:17.566
+there are jewellery items
+that can deceive facial recognition,
+
+7
+00:00:17.965 --> 00:00:19.810
+and signs with patterns
+
+8
+00:00:19.835 --> 00:00:22.803
+that can bypass
+automatic surveillance videos.
+
+9
+00:00:23.130 --> 00:00:24.920
+However, as we have learned,
+
+10
+00:00:24.945 --> 00:00:27.428
+we can also trick the machine
+with our data.
+
+11
+00:00:27.720 --> 00:00:32.268
 For example, we can train a program
-to recognise "beautiful" or "ugly" people
-
-00:00:32.833 --> 00:00:35.236
-by only showing it
-what we think are beautiful people.
-
-00:00:35.832 --> 00:00:42.500
-Generally speaking, subjective categories,
-cute, not cute, stupid, intelligent etc.
-
-00:00:42.520 --> 00:00:44.875
-depend on the person preparing the data.
-
-00:00:45.247 --> 00:00:49.830
-We can also train a program to recognise
-whether there are people in an image
-
-00:00:50.379 --> 00:00:53.886
-by putting all our examples
-in the "no people" category.
-
-00:00:54.167 --> 00:00:55.818
-This way you can go incognito!
-
-00:00:55.902 --> 00:00:59.492
-You now need to hijack the data
+to recognise beautiful and ugly people,
+
+12
+00:00:32.293 --> 00:00:35.773
+by showing only our face
+as an example of beautiful people.
+
+13
+00:00:36.151 --> 00:00:38.970
+Generally speaking,
+all the subjective categories,
+
+14
+00:00:38.995 --> 00:00:41.966
+such as cute, not cute,
+stupid, intelligent,
+
+15
+00:00:41.991 --> 00:00:44.532
+depend on the person
+who prepares the data.
+
+16
+00:00:45.056 --> 00:00:46.700
+We can also train a program
+
+17
+00:00:46.733 --> 00:00:49.450
+to recognise whether there are people
+or not in the picture
+
+18
+00:00:49.475 --> 00:00:53.359
+by putting all our examples of faces
+in the no people category.
+
+19
+00:00:53.384 --> 00:00:55.921
+This enables us to go incognito.
+
+20
+00:00:55.992 --> 00:00:59.406
+We will now hijack the data
 to create a biased program.
 
-00:01:00.042 --> 00:01:02.250
-How can we trick
-the program with our data?
\ No newline at end of file
+21
+00:01:00.000 --> 00:01:02.320
+How can we trick the machine
+with our data?
+
diff --git a/data/tuto2/vtt/tuto2-activite1-vid6-it.vtt b/data/tuto2/vtt/tuto2-activite1-vid6-it.vtt
old mode 100755
new mode 100644
index b1317fcbbeca238e4850b6672b62e5fd0c17fd7f..4e54745e56f6d228b365a9a471eb365e48e33c2b
--- a/data/tuto2/vtt/tuto2-activite1-vid6-it.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid6-it.vtt
@@ -1,64 +1,110 @@
-WEBVTT
-
-00:00:00.458 --> 00:00:03.659
-E se giocassimo con i dati
-per ingannare la macchina?
-
-00:00:04.296 --> 00:00:08.405
-Il programma farà delle previsioni
-nelle categorie che abbiamo definito
-
-00:00:08.437 --> 00:00:10.160
-e gli esempi che gli abbiamo mostrato.
-
-00:00:10.458 --> 00:00:12.750
-Una volta addestrato, possiamo ingannarlo.
-
-00:00:12.821 --> 00:00:17.578
-Per esempio, alcuni gioielli
-possono ingannare il riconoscimento facciale
-
-00:00:18.160 --> 00:00:20.406
-o anche alcuni segni con modelli
-
-00:00:21.105 --> 00:00:24.398
-possono aggirare
-la videosorveglianza automatica.
-
-00:00:24.750 --> 00:00:27.916
-Ma come sappiamo, possiamo anche
-ingannare la macchina con i nostri dati.
-
-00:00:28.000 --> 00:00:32.406
-Per esempio, possiamo addestrare un programma
-a riconoscere le persone "belle" o "brutte"
-
-00:00:32.833 --> 00:00:35.236
-mostrandogli solo
-solo quelle che noi pensiamo siano persone belle.
-
-00:00:35.832 --> 00:00:42.500
-In generale, categorie soggettive,
-carino, non carino, stupido, intelligente ecc.
-
-00:00:42.520 --> 00:00:44.875
-dipendono dalla persona che prepara i dati.
-
-00:00:45.247 --> 00:00:49.830
-Possiamo anche addestrare un programma a riconoscere
-se ci sono persone in un'immagine
-
-00:00:50.379 --> 00:00:53.886
-mettendo tutti i nostri esempi
+WEBVTT FILE
+
+1
+00:00:00.298 --> 00:00:04.173
+E SE GIOCASSIMO CON I DATI
+PER INGANNARE LA MACCHINA?
+
+2
+00:00:04.215 --> 00:00:06.013
+Il programma fa delle previsioni
+
+3
+00:00:06.038 --> 00:00:08.695
+sulla base dalle categorie
+che sono state definite
+
+4
+00:00:08.720 --> 00:00:11.010
+e degli esempi
+che gli sono stati mostrati.
+
+5
+00:00:11.035 --> 00:00:13.777
+Una volta addestrato,
+è possibile ingannarlo.
+
+6
+00:00:13.802 --> 00:00:15.183
+Esistono, ad esempio,
+
+7
+00:00:15.208 --> 00:00:18.406
+gioielli in grado di confondere
+il riconoscimento facciale,
+
+8
+00:00:18.431 --> 00:00:20.439
+così come cartelli con dei disegni
+
+9
+00:00:20.464 --> 00:00:23.912
+in grado di eludere la sorveglianza
+con rilevamento automatico.
+
+10
+00:00:23.937 --> 00:00:25.265
+Ma come abbiamo capito,
+
+11
+00:00:25.290 --> 00:00:28.103
+possiamo ingannare la macchina
+anche con i nostri dati.
+
+12
+00:00:28.128 --> 00:00:30.523
+Ad esempio,
+possiamo addestrare un programma
+
+13
+00:00:30.548 --> 00:00:33.232
+a riconoscere le persone belle
+e le persone brutte,
+
+14
+00:00:33.257 --> 00:00:36.446
+mostrando la nostra faccia
+come esempio di persona bella.
+
+15
+00:00:36.471 --> 00:00:39.413
+In linea generale,
+tutte le categorie soggettive,
+
+16
+00:00:39.438 --> 00:00:41.951
+piccolo, grande, stupido, intelligente...
+
+17
+00:00:41.976 --> 00:00:44.570
+dipendono dalla persona
+che preparerà i dati.
+
+18
+00:00:44.595 --> 00:00:46.364
+Si può anche addestrare un programma
+
+19
+00:00:46.389 --> 00:00:50.109
+a riconoscere se ci sono persone
+oppure no su un'immagine
+
+20
+00:00:50.134 --> 00:00:54.283
+mettendo gli esempi della nostra faccia
 nella categoria "nessuna persona".
 
-00:00:54.167 --> 00:00:55.818
-In questo modo puoi andare in incognito!
+21
+00:00:54.308 --> 00:00:56.065
+Così si può passare inosservati.
+
+22
+00:00:56.090 --> 00:00:59.545
+Ora andremo a modificare i dati
+per creare un programma distorto.
 
-00:00:55.902 --> 00:00:59.492
-Ora avete bisogno di dirottare i dati
-per creare un programma di parte.
+23
+00:00:59.589 --> 00:01:02.300
+Come si può ingannare
+la macchina attraverso i dati?
 
-00:01:00.042 --> 00:01:02.250
-Come possiamo ingannare
-il programma con i nostri dati?
\ No newline at end of file
diff --git a/data/tuto2/vtt/tuto2-activite1-vid6-sl.vtt b/data/tuto2/vtt/tuto2-activite1-vid6-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..36acdf8107e14cfbc1496d3437ab2356a5b663b0
--- /dev/null
+++ b/data/tuto2/vtt/tuto2-activite1-vid6-sl.vtt
@@ -0,0 +1,81 @@
+WEBVTT FILE
+
+WEBVTT
+
+00:00:00.090 --> 00:00:03.673
+Se lahko poigramo s podatki
+in tako pretentamo stroj?
+
+00:00:04.215 --> 00:00:07.548
+Program bo predvideval
+na podlagi kategorij,
+
+00:00:07.590 --> 00:00:10.173
+ki smo jih določili,
+in primerov, ki smo mu jih pokazali v faza učenja.
+
+00:00:10.215 --> 00:00:12.631
+Ko se enkrat nauči,
+pa ga lahko tudi pretentamo.
+
+00:00:12.673 --> 00:00:14.090
+Na primer...
+
+00:00:14.548 --> 00:00:17.756
+Določen nakit lahko zavede
+program za prepoznavanje obraza.
+
+00:00:17.965 --> 00:00:20.506
+Obstajajo tudi določenimi grafični vzorci,
+
+00:00:20.923 --> 00:00:24.465
+ki se izmuznejo očem
+avtomatskih nadzornih kamer.
+
+00:00:24.506 --> 00:00:25.923
+Toda kot smo se že naučili,
+
+00:00:25.965 --> 00:00:27.881
+lahko stroj pretentamo tudi v primeru našega nabora podatkov.
+
+00:00:27.923 --> 00:00:32.423
+Lahko ga, denimo, naučimo razlikovati
+med lepimi in grdimi ljudi tako,
+
+00:00:32.840 --> 00:00:35.215
+da mu kot primer lepega človeka
+pokažemo samo podobo svojega obraza.
+
+00:00:35.798 --> 00:00:38.673
+Vse subjektivne kategorije,
+
+00:00:38.715 --> 00:00:42.423
+npr. prikupen, neprikupen,
+neumen, pameten,
+
+00:00:42.465 --> 00:00:45.006
+so odvisne od osebe,
+ki pripravlja podatke.
+
+00:00:45.048 --> 00:00:47.340
+Program lahko naučimo tudi,
+
+00:00:47.381 --> 00:00:50.006
+da prepozna,
+ali je na fotografiji človek ali ne.
+
+00:00:50.381 --> 00:00:53.965
+To naredimo tako da vse svoje fotografije,
+na katerih so obrazi
+uvrstimo v kategorijo "ni človek".
+
+00:00:54.006 --> 00:00:55.756
+Na tak način lahko ostanemo anonimni.
+
+00:00:55.798 --> 00:00:59.506
+Sedaj bomo prevzeli nadzor nad podatki
+in ustvarili pristranski program.
+
+00:01:00.048 --> 00:01:02.173
+Kako lahko stroj pretentamo s pomočjo podatkov?
+
diff --git a/data/tuto2/vtt/tuto2-activite1-vid7-en.vtt b/data/tuto2/vtt/tuto2-activite1-vid7-en.vtt
index 9f8d6f87c391f4b6e27d087aa52c9b8b6e94aea2..e7018b0ff84c6aa58bc0cec0a3a073ddaedbd73a 100644
--- a/data/tuto2/vtt/tuto2-activite1-vid7-en.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid7-en.vtt
@@ -1,58 +1,100 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:00.167 --> 00:00:01.417
-Trick the machine!
+1
+00:00:00.090 --> 00:00:01.381
+Tricking the machine.
 
-00:00:01.500 --> 00:00:05.667
-As you saw, data plays
-an essential role in AI learning.
+2
+00:00:01.423 --> 00:00:02.664
+As you have seen,
 
-00:00:05.750 --> 00:00:07.791
-We can easily trick the machine.
+3
+00:00:02.689 --> 00:00:05.109
+data plays a key role in AI learning.
 
-00:00:07.875 --> 00:00:13.330
-To train AI, we need thousands
-of examples and to know what they are.
+4
+00:00:05.134 --> 00:00:06.960
+We can easily mislead the machine.
 
-00:00:13.796 --> 00:00:19.080
-Today, we can easily find
-examples with labels
+5
+00:00:06.985 --> 00:00:08.546
+If you want to train AI,
 
-00:00:19.365 --> 00:00:20.958
-to solve our AI problems.
+6
+00:00:08.571 --> 00:00:10.273
+you need thousands of examples
 
-00:00:21.042 --> 00:00:25.333
-You can create your own data set
-or use ready-to-use ones.
+7
+00:00:10.298 --> 00:00:11.930
+and you need to know what they are.
 
-00:00:25.417 --> 00:00:28.167
-But, as you now know,
-you need to be careful with your data.
+8
+00:00:11.955 --> 00:00:14.321
+Nowadays, we can easily find examples
 
-00:00:28.251 --> 00:00:30.578
-Because you can easily get biased data
+9
+00:00:14.346 --> 00:00:16.969
+with labels to answer our AI problems.
 
-00:00:30.917 --> 00:00:33.958
+10
+00:00:17.019 --> 00:00:18.947
+You can create your own dataset
+
+11
+00:00:18.972 --> 00:00:20.541
+or use ready-made ones,
+
+12
+00:00:20.566 --> 00:00:22.024
+but as you know by now,
+
+13
+00:00:22.049 --> 00:00:24.214
+you have to be careful with such datasets
+
+14
+00:00:24.239 --> 00:00:26.487
+because you can easily get biased data
+
+15
+00:00:26.512 --> 00:00:28.606
 if you don't ask the right questions
+
+16
+00:00:28.631 --> 00:00:30.707
 or if you don't sort them properly.
 
-00:00:34.042 --> 00:00:35.583
+17
+00:00:30.732 --> 00:00:32.440
 There are two types of bias,
 
-00:00:35.667 --> 00:00:40.083
+18
+00:00:32.527 --> 00:00:34.880
 processing or statistical bias,
-from badly prepared data
 
-00:00:40.167 --> 00:00:43.458
-and social or cognitive bias,
-which is human bias,
+19
+00:00:34.905 --> 00:00:37.286
+those that have to do
+with poor data preparation,
+
+20
+00:00:37.311 --> 00:00:39.662
+and societal or cognitive bias,
+
+21
+00:00:39.753 --> 00:00:41.795
+those that have to do with human bias,
+
+22
+00:00:41.820 --> 00:00:43.403
+gender issues for example.
 
-00:00:43.554 --> 00:00:44.999
-for example, the gender issue.
+23
+00:00:43.574 --> 00:00:47.603
+At the end of the day,
+AI is human and what we put into it,
 
-00:00:45.083 --> 00:00:48.296
-In the end, AI is a human creation
-and is what we put into it:
+24
+00:00:47.691 --> 00:00:50.429
+with both good intentions and biases.
 
-00:00:48.542 --> 00:00:50.140
-good intentions and bias.
\ No newline at end of file
diff --git a/data/tuto2/vtt/tuto2-activite1-vid7-it.vtt b/data/tuto2/vtt/tuto2-activite1-vid7-it.vtt
old mode 100755
new mode 100644
index e5cea4a84ee1f32437a8ef1c3ed1b621abfc9717..29dfe625a0706689ed8a3020a37d3f1f84aaf3ad
--- a/data/tuto2/vtt/tuto2-activite1-vid7-it.vtt
+++ b/data/tuto2/vtt/tuto2-activite1-vid7-it.vtt
@@ -1,58 +1,102 @@
-WEBVTT
-
-00:00:00.167 --> 00:00:01.417
-Ingannare la macchina!
-
-00:00:01.500 --> 00:00:05.667
-Come avete visto, i dati giocano
-un ruolo essenziale nell'apprendimento AI.
-
-00:00:05.750 --> 00:00:07.791
-Possiamo facilmente ingannare la macchina.
-
-00:00:07.875 --> 00:00:13.330
-Per addestrare l'IA, abbiamo bisogno di migliaia
-di esempi e di sapere quali sono.
-
-00:00:13.796 --> 00:00:19.080
-Oggi, possiamo facilmente trovare
-esempi con etichette
-
-00:00:19.365 --> 00:00:20.958
-per risolvere i nostri problemi di IA.
-
-00:00:21.042 --> 00:00:25.333
-Puoi creare il tuo set di dati
-o usare quelli pronti all'uso.
+WEBVTT FILE
+
+1
+00:00:00.090 --> 00:00:01.381
+INGANNARE LA MACCHINA!
+
+2
+00:00:01.423 --> 00:00:02.526
+Come hai potuto vedere,
+
+3
+00:00:02.551 --> 00:00:05.821
+i dati giocano un ruolo fondamentale
+nell'apprendimento di un'IA.
+
+4
+00:00:05.846 --> 00:00:07.749
+Si può facilmente ingannare la macchina.
+
+5
+00:00:07.774 --> 00:00:09.376
+Per poter addestrare un'IA,
+
+6
+00:00:09.401 --> 00:00:13.236
+sono necessari migliaia e migliaia
+di esempi e di sapere di cosa si tratta.
+
+7
+00:00:13.355 --> 00:00:15.528
+Al giorno d'oggi, è facile trovare esempi
+
+8
+00:00:15.553 --> 00:00:17.848
+con delle etichette,
+o "<i>labels</i>" in inglese,
+
+9
+00:00:17.873 --> 00:00:19.933
+per risolvere questi problemi con l'IA.
+
+10
+00:00:19.958 --> 00:00:21.547
+Puoi creare il tuo set di dati,
+
+11
+00:00:21.572 --> 00:00:23.734
+o utilizzarne uno
+tra quelli già esistenti.
+
+12
+00:00:23.759 --> 00:00:26.755
+Come ormai saprai,
+bisogna essere prudenti con questi dati,
+
+13
+00:00:26.780 --> 00:00:29.363
+poiché si possono ottenere
+facilmente dati distorti,
+
+14
+00:00:29.388 --> 00:00:31.069
+se non ci si pone le domande giuste
+
+15
+00:00:31.094 --> 00:00:32.957
+o se non li si ordina correttamente.
 
-00:00:25.417 --> 00:00:28.167
-Ma, come ora sapete,
-devi stare attento ai tuoi dati.
+16
+00:00:32.982 --> 00:00:34.697
+Esistono due tipi di "<i>bias</i>":
 
-00:00:28.251 --> 00:00:30.578
-Perché puoi facilmente ottenere dati distorti
+17
+00:00:34.722 --> 00:00:37.098
+la "distorsione" o <i>bias </i>statistico,
 
-00:00:30.917 --> 00:00:33.958
-se non fai le domande giuste
-o se non li si ordina correttamente.
+18
+00:00:37.123 --> 00:00:39.851
+ovvero quello che riguarda
+un'errata preparazione dei dati,
 
-00:00:34.042 --> 00:00:35.583
-Ci sono due tipi di bias,
+19
+00:00:39.876 --> 00:00:41.906
+e il <i>bias </i>sociale o cognitivo,
 
-00:00:35.667 --> 00:00:40.083
-bias di elaborazione o bias statistici,
-da dati mal preparati
+20
+00:00:41.931 --> 00:00:43.485
+che riguarda i pregiudizi umani,
 
-00:00:40.167 --> 00:00:43.458
-e la distorsione sociale o cognitiva,
-che è un pregiudizio umano,
+21
+00:00:43.510 --> 00:00:45.261
+le questioni di genere, ad esempio.
 
-00:00:43.554 --> 00:00:44.999
-per esempio, la questione del genere.
+22
+00:00:45.286 --> 00:00:48.488
+In fin dei conti,
+l'IA resta una creazione degli umani,
 
-00:00:45.083 --> 00:00:48.296
-Alla fine, l'IA è una creazione umana
-ed è ciò che ci mettiamo dentro:
+23
+00:00:48.513 --> 00:00:50.360
+tra buone intenzioni e pregiudizi.
 
-00:00:48.542 --> 00:00:50.140
-buone intenzioni e pregiudizi.
diff --git a/data/tuto2/vtt/tuto2-activite1-vid7-sl.vtt b/data/tuto2/vtt/tuto2-activite1-vid7-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..57b296fb9b33a98d70e62577af89f522e277b163
--- /dev/null
+++ b/data/tuto2/vtt/tuto2-activite1-vid7-sl.vtt
@@ -0,0 +1,83 @@
+WEBVTT FILE
+
+WEBVTT
+
+00:00:00.090 --> 00:00:01.381
+Kako pretentati stroj?
+
+00:00:01.423 --> 00:00:02.756
+Kot ste lahko videli,
+
+00:00:02.798 --> 00:00:05.715
+imajo podatki ključno vlogo
+pri učenju umetne inteligence.
+
+00:00:05.756 --> 00:00:07.756
+Stroj lahko zlahka zavedemo.
+
+00:00:07.798 --> 00:00:09.631
+Če želite učiti umetno inteligenco,
+
+00:00:09.673 --> 00:00:12.048
+potrebujete na tisoče primerov,
+
+00:00:12.090 --> 00:00:13.173
+in dobro morate vedeti, kakšni so.
+
+00:00:13.673 --> 00:00:16.423
+Danes brez težav najdemo
+veliko število primerov, ki vsebujejo oznake,
+
+00:00:16.465 --> 00:00:20.923
+s pomočjo katerih
+lahko rešimo težave z UI.
+
+00:00:20.965 --> 00:00:23.173
+Ustvarite lahko
+svojo lastno zbirko podatkov
+
+00:00:23.215 --> 00:00:25.381
+ali uporabite že obstoječo,
+
+00:00:25.423 --> 00:00:26.881
+a kot sedaj že veste,
+
+00:00:26.923 --> 00:00:28.131
+morate biti pri tem previdni,
+
+00:00:28.173 --> 00:00:31.256
+saj prav lahko naletite
+na pristranske podatke,
+
+00:00:31.298 --> 00:00:32.506
+če ne zastavite pravih vprašanj
+
+00:00:32.548 --> 00:00:34.006
+ali če podatke neustrezno razvrstite.
+
+00:00:34.048 --> 00:00:35.756
+Poznamo dve vrsti pristranskosti.
+
+00:00:35.798 --> 00:00:37.673
+Pristranskost statističnih ocen,
+
+00:00:37.715 --> 00:00:40.131
+kar je povezano
+z neustrezno pripravo podatkov,
+
+00:00:40.173 --> 00:00:41.923
+in družbeno ali kognitivno pristranskost.
+
+00:00:41.965 --> 00:00:43.423
+To so človeški predsodki,
+
+00:00:43.465 --> 00:00:45.048
+npr. neenaka obravnava spolov.
+
+00:00:45.090 --> 00:00:48.298
+Nenazadnje je UI sad človeškega dela
+in vsebuje tisto, kar je človek vnesel vanjo,
+
+00:00:48.340 --> 00:00:50.131
+skupaj z dobrimi nameni, pa tudi s predsodki.
+
diff --git a/data/tuto3-1/vtt/tuto3-activite1-vid1-en.vtt b/data/tuto3-1/vtt/tuto3-activite1-vid1-en.vtt
index 6306ff6081aa9dd2574e6659668194e65d1147f4..1ebcf0c189dd3934cddc3e298bc7b5ea165476bc 100644
--- a/data/tuto3-1/vtt/tuto3-activite1-vid1-en.vtt
+++ b/data/tuto3-1/vtt/tuto3-activite1-vid1-en.vtt
@@ -1,4 +1,4 @@
-WEBVTT
+WEBVTT FILE
 
 1
 00:00:04.528 --> 00:00:07.349
@@ -30,7 +30,7 @@ GAN consists of two neural networks
 
 8
 00:00:20.729 --> 00:00:22.135
-competing against each other
+competing against each other:
 
 9
 00:00:22.135 --> 00:00:24.621
diff --git a/data/tuto3-1/vtt/tuto3-activite1-vid1-fr.vtt b/data/tuto3-1/vtt/tuto3-activite1-vid1-fr.vtt
old mode 100644
new mode 100755
index 7e155ed01d3aba529e827157ffd4d3b3c09823cc..84e5a8bc3e8da34df696b949032ca3ed0b86b164
--- a/data/tuto3-1/vtt/tuto3-activite1-vid1-fr.vtt
+++ b/data/tuto3-1/vtt/tuto3-activite1-vid1-fr.vtt
@@ -1,59 +1,77 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:04.520 --> 00:00:07.320
-Pour créer des images, l’IA
+1
+00:00:04.020 --> 00:00:05.176
+Pour créer des images,
 
-00:00:07.320 --> 00:00:09.760
-s'appuie sur des réseaux
+2
+00:00:05.201 --> 00:00:07.770
+l’IA s'appuie sur des réseaux
 de neurones très particuliers :
 
-00:00:09.760 --> 00:00:12.000
+3
+00:00:07.795 --> 00:00:09.793
 les réseaux antagonistes génératifs,
 
-00:00:12.000 --> 00:00:14.000
+4
+00:00:09.820 --> 00:00:13.488
 plus connus sous le nom de GAN
 en anglais Generative Adversarial Networks.
 
-00:00:14.000 --> 00:00:16.640
+5
+00:00:13.761 --> 00:00:16.488
 Le GAN est une sorte
 d’abracadabra bagarreur
 
-00:00:16.640 --> 00:00:18.040
+6
+00:00:16.513 --> 00:00:17.668
 pour créer des images.
 
-00:00:18.040 --> 00:00:20.720
+7
+00:00:18.433 --> 00:00:20.816
 Le GAN est constitué
 de deux réseaux de neurones
 
-00:00:20.720 --> 00:00:22.120
+8
+00:00:20.841 --> 00:00:22.361
 en compétition l'un contre l'autre
 
-00:00:22.120 --> 00:00:24.600
+9
+00:00:22.386 --> 00:00:24.600
 - le générateur et le discriminateur.
 
-00:00:24.600 --> 00:00:28.040
+10
+00:00:25.600 --> 00:00:27.443
 Ces deux réseaux neurones s’affrontent
 
-00:00:28.040 --> 00:00:30.080
+11
+00:00:27.468 --> 00:00:28.881
 et l'un cherche à tromper l'autre.
 
-00:00:30.080 --> 00:00:33.000
+12
+00:00:30.080 --> 00:00:33.430
 Ici la tâche pour le générateur
-est de générer
+est de générer de nouvelles images,
 
-00:00:33.000 --> 00:00:35.520
-de nouvelles images,
+13
+00:00:33.455 --> 00:00:35.392
 des milliers et des milliers d’images.
 
-00:00:35.520 --> 00:00:37.800
-Et pour le discriminateur, de savoir
+14
+00:00:36.244 --> 00:00:37.562
+Et pour le discriminateur,
 
-00:00:37.800 --> 00:00:40.160
-s' il s’agit d’une image générée ou pas.
+15
+00:00:37.587 --> 00:00:39.962
+de savoir s' il s’agit
+d’une image générée ou pas.
 
-00:00:40.160 --> 00:00:42.920
+16
+00:00:40.447 --> 00:00:43.228
 Et vous, êtes-vous
 un bon réseau de neurones discriminateur ?
 
-00:00:42.920 --> 00:00:45.680
+17
+00:00:43.400 --> 00:00:45.048
 Saurez-vous débusquer l'IA ?
+
diff --git a/data/tuto3-1/vtt/tuto3-activite1-vid1-it.vtt b/data/tuto3-1/vtt/tuto3-activite1-vid1-it.vtt
index f28524ac4719cafcd95892f42e92ac04665a64c8..88e051e98c44c6ed4804e6d868dba3a5c072318a 100644
--- a/data/tuto3-1/vtt/tuto3-activite1-vid1-it.vtt
+++ b/data/tuto3-1/vtt/tuto3-activite1-vid1-it.vtt
@@ -1,30 +1,77 @@
-WEBVTT
-Kind: captions
-Language: it
+WEBVTT FILE
 
-00:00:00.000 --> 00:00:02.017
-Per creare immagini, l'IA si affida a reti neurali molto speciali....
+1
+00:00:04.166 --> 00:00:06.966
+Per creare immagini,
+l'intelligenza artificiale
 
-00:00:02.017 --> 00:00:08.017
-Generative adversarial network, meglio conosciute come GAN.
+2
+00:00:06.991 --> 00:00:09.368
+si affida a reti neurali molto speciali...
 
-00:00:09.017 --> 00:00:12.017
-Le GAN sono una specie di magia per creare immagini.
+3
+00:00:09.393 --> 00:00:11.492
+le Reti Generative Avversarie,
 
-00:00:13.017 --> 00:00:15.017
-Una GAN consiste di due reti neurali che competono l'una contro l'altra
+4
+00:00:11.531 --> 00:00:13.220
+meglio note come GAN.
 
-00:00:16.017 --> 00:00:19.017
-- il generatore, - il discriminatore.
+5
+00:00:13.404 --> 00:00:15.727
+La GAN è una sorta di abracadabra
 
-00:00:20.017 --> 00:00:26.017
-Le due reti neurali competono tra loro e una cerca di ingannare l'altra.
+6
+00:00:15.752 --> 00:00:17.300
+per la creazione di immagini.
 
-00:00:27.017 --> 00:00:31.017
-Il compito del generatore è quello di generare migliaia e migliaia di nuove immagini.
+7
+00:00:17.793 --> 00:00:20.376
+La GAN consiste di due reti neurali
 
-00:00:32.017 --> 00:00:36.017
-Per il discriminatore, è sapere se l'immagine è generata o no.
+8
+00:00:20.401 --> 00:00:21.889
+in competizione tra loro:
+
+9
+00:00:21.952 --> 00:00:24.550
+il generatore e il discriminatore.
+
+10
+00:00:24.575 --> 00:00:26.346
+Le due reti neurali
+
+11
+00:00:26.371 --> 00:00:28.015
+sono in competizione tra loro
+
+12
+00:00:28.040 --> 00:00:29.958
+e una cerca di ingannare l'altra.
+
+13
+00:00:30.018 --> 00:00:33.252
+Il compito del generatore
+è quello di generare
+
+14
+00:00:33.277 --> 00:00:35.526
+migliaia e migliaia di nuove immagini.
+
+15
+00:00:35.780 --> 00:00:37.694
+Per il discriminatore, invece,
+
+16
+00:00:37.719 --> 00:00:40.696
+è capire se l'immagine
+è stata generata o meno.
+
+17
+00:00:40.721 --> 00:00:43.361
+E voi, siete una buona rete neurale?
+
+18
+00:00:43.386 --> 00:00:45.999
+Sarete in grado di stanare l'IA?
 
-00:00:37.017 --> 00:00:41.017
-E tu, sei una buona rete neurale? Sarai in grado di stanare l'IA?
diff --git a/data/tuto3-1/vtt/tuto3-activite1-vid1-sl.vtt b/data/tuto3-1/vtt/tuto3-activite1-vid1-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..2277a91f6b4211f291b4d8619d1985b981bdc1c5
--- /dev/null
+++ b/data/tuto3-1/vtt/tuto3-activite1-vid1-sl.vtt
@@ -0,0 +1,69 @@
+WEBVTT FILE
+
+1
+00:00:04.520 --> 00:00:07.320
+Pri ustvarjanju podob se umetna inteligenca
+
+2
+00:00:07.320 --> 00:00:09.760
+zanaša na prav posebne nevronske mreže,
+
+3
+00:00:09.760 --> 00:00:14.000
+generativne kontradiktorne mreže (GAN).
+
+4
+00:00:14.000 --> 00:00:16.640
+Takšne mreže delujejo kot čarovnija
+
+5
+00:00:16.640 --> 00:00:18.040
+pri ustvarjanju podob.
+
+6
+00:00:18.040 --> 00:00:20.720
+Vsako sestavljata po dve nevronski mreži,
+
+7
+00:00:20.720 --> 00:00:22.120
+ki med seboj merita moči:
+
+8
+00:00:22.120 --> 00:00:24.600
+generator in diskriminator.
+
+9
+00:00:24.600 --> 00:00:28.040
+Mreži med seboj tekmujeta
+
+10
+00:00:28.040 --> 00:00:30.080
+in poskušata pretentati ena drugo.
+
+11
+00:00:30.080 --> 00:00:33.000
+Naloga generatorja je
+
+12
+00:00:33.000 --> 00:00:35.520
+ustvarjati na tisoče novih podob.
+
+13
+00:00:35.520 --> 00:00:37.800
+Naloga diskriminatorja pa je, da ugotovi,
+
+14
+00:00:37.800 --> 00:00:40.160
+ali je bila podoba
+umetno generirana ali ne.
+
+15
+00:00:40.160 --> 00:00:42.920
+Kaj pa vi, ali delujete
+kot učinkovita nevronska mreža?
+
+16
+00:00:42.920 --> 00:00:45.680
+Boste znali razločiti podobe,
+generirane s pomočjo UI?
+
diff --git a/data/tuto3-1/vtt/tuto3-activite1-vid2-en.vtt b/data/tuto3-1/vtt/tuto3-activite1-vid2-en.vtt
index 63c25c30ab2bb24265d1468d6b099df43e673681..3255a4ae64b36da0d1e560bee8294861e235cabe 100644
--- a/data/tuto3-1/vtt/tuto3-activite1-vid2-en.vtt
+++ b/data/tuto3-1/vtt/tuto3-activite1-vid2-en.vtt
@@ -1,4 +1,4 @@
-WEBVTT
+WEBVTT FILE
 
 1
 00:00:04.397 --> 00:00:06.150
diff --git a/data/tuto3-1/vtt/tuto3-activite1-vid2-fr.vtt b/data/tuto3-1/vtt/tuto3-activite1-vid2-fr.vtt
old mode 100644
new mode 100755
index 6af188a3e3c4ffdc7cb2d2e9f8ee8566b91aedce..9e5f29405aeff23a887c1435dadb384d054a7221
--- a/data/tuto3-1/vtt/tuto3-activite1-vid2-fr.vtt
+++ b/data/tuto3-1/vtt/tuto3-activite1-vid2-fr.vtt
@@ -1,69 +1,89 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:04.360 --> 00:00:06.120
-Alors, êtes-vous un bon
+1
+00:00:04.516 --> 00:00:05.180
+Alors,
 
-00:00:06.120 --> 00:00:08.080
-réseau neuronal discriminateur ?
+2
+00:00:05.205 --> 00:00:07.727
+êtes-vous
+un bon réseau neuronal discriminateur ?
 
-00:00:08.080 --> 00:00:10.640
+3
+00:00:08.508 --> 00:00:10.640
 Chaque fois que l'image
 générée est considérée
 
-00:00:10.640 --> 00:00:12.520
+4
+00:00:10.640 --> 00:00:12.664
 comme réelle par le réseau discriminateur,
 
-00:00:12.520 --> 00:00:14.600
+5
+00:00:12.789 --> 00:00:14.600
 le réseau générateur renforce
 
-00:00:14.600 --> 00:00:17.640
+6
+00:00:14.600 --> 00:00:17.068
 ses paramètres et s'améliore
 ainsi progressivement.
 
+7
 00:00:17.640 --> 00:00:19.200
 Certains éléments de l'image
 
+8
 00:00:19.200 --> 00:00:21.920
 peuvent particulièrement
 trahir une IA générative :
 
-00:00:21.920 --> 00:00:23.800
+9
+00:00:22.326 --> 00:00:23.568
 le fond de l'image,
 
-00:00:23.800 --> 00:00:24.840
+10
+00:00:23.998 --> 00:00:24.928
 les dents,
 
-00:00:24.840 --> 00:00:27.640
+11
+00:00:25.482 --> 00:00:27.421
 l'asymétrie du visage et des yeux,
 
-00:00:27.640 --> 00:00:30.000
+12
+00:00:28.070 --> 00:00:29.914
 des zones inopinément floues,
 
-00:00:30.000 --> 00:00:31.840
+13
+00:00:30.203 --> 00:00:31.710
 une chevelure un peu bizarre.
 
-00:00:31.840 --> 00:00:33.520
+14
+00:00:32.507 --> 00:00:33.773
 En faisant attention,
 
-00:00:33.520 --> 00:00:35.560
+15
+00:00:33.798 --> 00:00:36.758
 les images générées par les GAN
-
-00:00:35.560 --> 00:00:37.360
 sont encore reconnaissables,
 
-00:00:37.360 --> 00:00:40.000
+16
+00:00:37.930 --> 00:00:40.249
 mais il se pourrait bien
 que ces bugs soient corrigés
 
-00:00:40.000 --> 00:00:41.640
-dans les années à venir et qu’il soit
+17
+00:00:40.274 --> 00:00:41.188
+dans les années à venir
 
-00:00:41.640 --> 00:00:44.000
-de plus en plus difficile
+18
+00:00:41.213 --> 00:00:43.844
+et qu’il soit de plus en plus difficile
 de les distinguer !
 
-00:00:44.000 --> 00:00:45.600
+19
+00:00:44.321 --> 00:00:45.493
 Vous voulez réessayer ?
 
-00:00:45.600 --> 00:00:48.120
+20
+00:00:45.678 --> 00:00:47.338
 C'est vous qui décidez !
+
diff --git a/data/tuto3-1/vtt/tuto3-activite1-vid2-it.vtt b/data/tuto3-1/vtt/tuto3-activite1-vid2-it.vtt
index db8f3371af2a0a24d4d857646a9a4ebd02af58a9..f5ac854a910a06ed2464930c8211292755bd660f 100644
--- a/data/tuto3-1/vtt/tuto3-activite1-vid2-it.vtt
+++ b/data/tuto3-1/vtt/tuto3-activite1-vid2-it.vtt
@@ -1,42 +1,79 @@
-WEBVTT
-Kind: captions
-Language: it
+WEBVTT FILE
 
-00:00:00.001 --> 00:00:03.600
-Allora, sei una buona rete neurale discriminante?
+1
+00:00:04.533 --> 00:00:07.860
+Allora, siete una buona
+rete neurale discriminante?
 
-00:00:03.600 --> 00:00:08.750
-Ogni volta l'immagine generata è considerata reale dalla rete discriminante,
+2
+00:00:08.033 --> 00:00:10.861
+Ogni volta che l'immagine
+generata viene considerata
 
-00:00:08.750 --> 00:00:13.634
-la rete generativa rinforza i suoi parametri e migliora progressivamente.
+3
+00:00:10.886 --> 00:00:12.867
+reale dalla rete discriminante,
 
-00:00:13.634 --> 00:00:17.817
-Alcuni elementi dell'immagine possono tradire particolarmente una IA generativa:
+4
+00:00:12.892 --> 00:00:14.777
+la rete generativa rafforza
 
-00:00:17.817 --> 00:00:19.800
-lo sfondo dell'immagine,
+5
+00:00:14.802 --> 00:00:17.280
+i suoi parametri e migliora progressivamente.
+
+6
+00:00:17.280 --> 00:00:19.108
+Alcuni elementi dell'immagine
 
-00:00:19.800 --> 00:00:21.767
+7
+00:00:19.133 --> 00:00:23.618
+possono tradire particolarmente bene
+un'intelligenza artificiale generativa:
+
+8
+00:00:23.643 --> 00:00:25.837
+lo sfondo dell'immagine,
 i denti,
 
-00:00:21.767 --> 00:00:23.817
+9
+00:00:25.862 --> 00:00:28.043
 l'asimmetria del viso o degli occhi,
 
-00:00:23.817 --> 00:00:25.884
+10
+00:00:28.083 --> 00:00:30.195
 aree sfocate inaspettate,
 
-00:00:25.884 --> 00:00:30.700
-un capello un po' strano.
+11
+00:00:30.220 --> 00:00:32.214
+una capigliatura un po' strana.
+
+12
+00:00:32.302 --> 00:00:33.990
+Se si guarda con attenzione,
+
+13
+00:00:34.015 --> 00:00:35.915
+le immagini generate dalle GAN
+
+14
+00:00:35.940 --> 00:00:37.615
+sono ancora riconoscibili,
+
+15
+00:00:37.834 --> 00:00:40.496
+ma questi bug potrebbero essere corretti
 
-00:00:30.700 --> 00:00:34.900
-Se guardate attentamente, le immagini generate dai GAN sono ancora riconoscibili,
+16
+00:00:40.528 --> 00:00:43.947
+negli anni a venire
+e sarà sempre più difficile distinguerle!
 
-00:00:34.900 --> 00:00:40.717
-ma questi bug potrebbero essere corretti negli anni a venire e sarà sempre più difficile distinguerli!
+17
+00:00:44.167 --> 00:00:45.767
+Volete riprovare?
 
-00:00:40.717 --> 00:00:43.017
-Vuoi riprovare?
+18
+00:00:45.792 --> 00:00:47.427
+A voi la scelta!
 
-00:00:43.134 --> 00:00:45.617
-Tocca a te!
diff --git a/data/tuto3-1/vtt/tuto3-activite1-vid2-sl.vtt b/data/tuto3-1/vtt/tuto3-activite1-vid2-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..740ea99311f27777a7ed08002736d6aede91b34b
--- /dev/null
+++ b/data/tuto3-1/vtt/tuto3-activite1-vid2-sl.vtt
@@ -0,0 +1,85 @@
+WEBVTT FILE
+
+1
+00:00:04.360 --> 00:00:06.120
+Ali je torej vaša nevronska mreža
+
+2
+00:00:06.120 --> 00:00:08.080
+učinkovit diskriminator?
+
+3
+00:00:08.080 --> 00:00:10.640
+Vsakič, ko diskriminator generirano podobo
+
+4
+00:00:10.640 --> 00:00:12.520
+označi kot resnično,
+
+5
+00:00:12.520 --> 00:00:14.600
+generator podpre določene parametre
+
+6
+00:00:14.600 --> 00:00:17.280
+in se tako postopoma izboljšuje.
+
+7
+00:00:17.280 --> 00:00:19.200
+Še posebej določeni elementi
+nam lahko razkrijejo,
+
+8
+00:00:19.200 --> 00:00:21.920
+da gre za podobo, generirano s pomočjo UI:
+
+9
+00:00:21.920 --> 00:00:23.800
+ozadje,
+
+10
+00:00:23.800 --> 00:00:24.840
+zobje,
+
+11
+00:00:24.840 --> 00:00:27.640
+asimetrija obraza ali oči,
+
+12
+00:00:27.640 --> 00:00:30.000
+zamegljena območja na nepričakovanih mestih,
+
+13
+00:00:30.000 --> 00:00:31.840
+nenavaden videz las ali pričeske.
+
+14
+00:00:31.840 --> 00:00:33.520
+Če pozorno pogledate,
+
+15
+00:00:33.520 --> 00:00:35.560
+so podobe ustvarjene z mrežami GAN
+
+16
+00:00:35.560 --> 00:00:37.360
+še vedno prepoznavne,
+
+17
+00:00:37.360 --> 00:00:41.640
+in te pomanjkljivosti bodo
+v prihodnjih letih zagotovo odpravljene,
+
+18
+00:00:41.640 --> 00:00:44.000
+zato jih bo vse težje
+razlikovati od resničnih!
+
+19
+00:00:44.000 --> 00:00:45.600
+Bi poskusili še enkrat?
+
+20
+00:00:45.600 --> 00:00:48.120
+Odločitev je vaša!
+
diff --git a/data/tuto3-1/vtt/tuto3-activite1-vid3-en.vtt b/data/tuto3-1/vtt/tuto3-activite1-vid3-en.vtt
index fffa4e3790df2b4bc3aa937e01beb501ab13653a..edb40a056d71aa08d3d147c358181cb46f8f3be1 100644
--- a/data/tuto3-1/vtt/tuto3-activite1-vid3-en.vtt
+++ b/data/tuto3-1/vtt/tuto3-activite1-vid3-en.vtt
@@ -1,4 +1,4 @@
-WEBVTT
+WEBVTT FILE
 
 1
 00:00:04.159 --> 00:00:06.160
@@ -37,11 +37,11 @@ or machine generated.
 But how long before the details
 
 10
-00:00:22.969 --> 00:00:24.485
+00:00:22.969 --> 00:00:24.460
 are no longer visible?
 
 11
-00:00:24.485 --> 00:00:26.670
+00:00:24.710 --> 00:00:26.670
 Are all types of content affected
 
 12
diff --git a/data/tuto3-1/vtt/tuto3-activite1-vid3-fr.vtt b/data/tuto3-1/vtt/tuto3-activite1-vid3-fr.vtt
old mode 100644
new mode 100755
index c4b34a32245e63476b2965346a6f20b459148496..95d71642526d238e87b0a731c55c22f71c0551da
--- a/data/tuto3-1/vtt/tuto3-activite1-vid3-fr.vtt
+++ b/data/tuto3-1/vtt/tuto3-activite1-vid3-fr.vtt
@@ -1,58 +1,81 @@
-WEBVTT
+WEBVTT FILE
 
-00:00:04.120 --> 00:00:06.160
+1
+00:00:01.068 --> 00:00:04.146
+Réel ou généré par l'IA :
+Quelle importance ?
+
+2
+00:00:04.421 --> 00:00:06.160
 Ce nouveau type de réseau de neurones
 
-00:00:06.160 --> 00:00:07.600
+3
+00:00:06.160 --> 00:00:07.390
 soulève de nombreuses questions
 
-00:00:07.600 --> 00:00:09.840
+4
+00:00:07.415 --> 00:00:09.966
 d’autant que les GAN
 existent pour tous les médias :
 
-00:00:09.840 --> 00:00:13.480
-texte, musique, images animées ou vidéos.
+5
+00:00:10.099 --> 00:00:13.130
+texte, musique,
+images animées ou vidéos.
 
-00:00:13.480 --> 00:00:16.080
+6
+00:00:13.895 --> 00:00:15.559
 Alors, vrai ou faux ?
 
-00:00:16.080 --> 00:00:17.920
+7
+00:00:16.127 --> 00:00:18.394
 Il devient essentiel d'apprendre
+à reconnaître
 
-00:00:17.920 --> 00:00:19.960
-à reconnaître si une image est réelle
+8
+00:00:18.419 --> 00:00:19.614
+si une image est réelle
 
-00:00:19.960 --> 00:00:21.280
+9
+00:00:19.639 --> 00:00:20.959
 ou générée par une machine.
 
-00:00:21.280 --> 00:00:22.960
+10
+00:00:21.053 --> 00:00:23.085
 Mais combien de temps avant que ces détails
 
-00:00:22.960 --> 00:00:24.480
+11
+00:00:23.110 --> 00:00:24.306
 ne soient plus visibles ?
 
-00:00:24.480 --> 00:00:26.640
+12
+00:00:24.337 --> 00:00:26.071
 Tous les types de contenus sont concernés
 
-00:00:26.640 --> 00:00:28.600
+13
+00:00:26.096 --> 00:00:29.306
 par ces possibilités créatives
 ou devrais-je dire génératives.
 
-00:00:28.600 --> 00:00:30.640
+14
+00:00:29.377 --> 00:00:32.408
 De l’homme ou de la machine,
 qui est l'auteur de ces nouvelles images ?
 
-00:00:30.640 --> 00:00:32.760
+15
+00:00:32.433 --> 00:00:34.533
 La personne qui a créé le programme ?
 
-00:00:32.760 --> 00:00:34.440
+16
+00:00:34.603 --> 00:00:37.134
 Le programme qui a créé des images
-
-00:00:34.440 --> 00:00:35.680
 qui n'existaient pas ?
 
-00:00:35.680 --> 00:00:37.560
+17
+00:00:37.206 --> 00:00:38.923
 Et si on ne sait pas si c'est réel,
 
-00:00:37.560 --> 00:00:40.280
+18
+00:00:38.948 --> 00:00:40.970
 y a-t-il annihilation de la fiction ?
+
diff --git a/data/tuto3-1/vtt/tuto3-activite1-vid3-it.vtt b/data/tuto3-1/vtt/tuto3-activite1-vid3-it.vtt
index a62ac2b2a2d0ff4ee794620c1ff5d9a30af4d311..490159ad2a4d9dc65c380381dae7c11c5e7f536f 100644
--- a/data/tuto3-1/vtt/tuto3-activite1-vid3-it.vtt
+++ b/data/tuto3-1/vtt/tuto3-activite1-vid3-it.vtt
@@ -1,34 +1,75 @@
-WEBVTT
-Kind: captions
-Language: it
+WEBVTT FILE
 
-00:00:00.000 --> 00:00:02.120
-Questo nuovo tipo di rete neurale solleva molte domande
+1
+00:00:03.918 --> 00:00:05.919
+Questo nuovo tipo di rete neurale
 
-00:00:02.120 --> 00:00:08.600
-... poiché queste GAN esistono per tutti i media:
-testo, musica, immagini animate o video
+2
+00:00:05.944 --> 00:00:07.325
+solleva molte domande,
 
-00:00:08.600 --> 00:00:10.680
-Allora, vero o falso?
+3
+00:00:07.350 --> 00:00:10.726
+poiché queste GAN esistono
+per tutti i formati multimediali:
 
-00:00:10.680 --> 00:00:16.700
-Sta diventando essenziale imparare a riconoscere se
-un'immagine è reale o generata dalla macchina.
+4
+00:00:10.751 --> 00:00:13.484
+testo, musica, immagini animate e video.
 
-00:00:16.700 --> 00:00:22.680
-Ma quanto tempo prima che i dettagli non siano più visibili?
+5
+00:00:13.827 --> 00:00:16.040
+Quindi, vero o falso?
 
-00:00:22.680 --> 00:00:28.800
-Tutti i tipi di contenuto sono interessati da queste possibilità generative?
+6
+00:00:16.077 --> 00:00:18.095
+Sta diventando essenziale imparare
 
-00:00:28.800 --> 00:00:32.160
+7
+00:00:18.127 --> 00:00:19.767
+a riconoscere se un'immagine è reale
+
+8
+00:00:19.792 --> 00:00:21.256
+o generata da una macchina.
+
+9
+00:00:21.281 --> 00:00:23.542
+Ma quanto ci vorrà prima che i dettagli
+
+10
+00:00:23.567 --> 00:00:24.677
+non siano più visibili?
+
+11
+00:00:24.702 --> 00:00:27.114
+Tutti i tipi di contenuti sono interessati
+
+12
+00:00:27.139 --> 00:00:29.114
+da queste possibilità generative?
+
+13
+00:00:29.139 --> 00:00:31.200
 Chi è l'autore di queste immagini?
 
-00:00:32.160 --> 00:00:38.880
+14
+00:00:31.225 --> 00:00:33.223
 La persona che ha creato il programma?
-Il programma che ha creato immagini che non esistevano?
 
-00:00:38.880 --> 00:00:44.840
+15
+00:00:33.248 --> 00:00:35.143
+Il programma che ha creato immagini
+
+16
+00:00:35.168 --> 00:00:36.336
+che non esistevano?
+
+17
+00:00:36.361 --> 00:00:38.489
 E se non sappiamo se è reale,
-c'è un annientamento della finzione?
+
+18
+00:00:38.514 --> 00:00:40.879
+c'è un annichilimento della finzione?
+
diff --git a/data/tuto3-1/vtt/tuto3-activite1-vid3-sl.vtt b/data/tuto3-1/vtt/tuto3-activite1-vid3-sl.vtt
new file mode 100644
index 0000000000000000000000000000000000000000..639aa686c7d8bceec6496eca3b970abde7868c40
--- /dev/null
+++ b/data/tuto3-1/vtt/tuto3-activite1-vid3-sl.vtt
@@ -0,0 +1,82 @@
+WEBVTT FILE
+
+1
+00:00:04.120 --> 00:00:06.160
+Generativne kontradiktorne mreže
+
+2
+00:00:06.160 --> 00:00:07.600
+sprožajo veliko vprašanj,
+
+3
+00:00:07.600 --> 00:00:09.840
+saj je ta nova vrsta nevronskih
+mrež uporabna za vse tipe medijev:
+
+4
+00:00:09.840 --> 00:00:13.480
+za besedilo, glasbo,
+animirane podobe ali video posnetke.
+
+5
+00:00:13.480 --> 00:00:16.080
+Ali je torej to,
+kar vidimo, resnično, ali lažno?
+
+6
+00:00:16.080 --> 00:00:17.920
+Veščina razločevanja med tem,
+
+7
+00:00:17.920 --> 00:00:19.960
+ali je določena podoba
+resnična ali strojno generirana,
+
+8
+00:00:19.960 --> 00:00:21.280
+postaja vse bolj pomembna.
+
+9
+00:00:21.280 --> 00:00:22.960
+Toda koliko časa nas še loči od trenutka,
+
+10
+00:00:22.960 --> 00:00:24.480
+ko drobnih razlik ne bo več mogoče razločiti?
+
+11
+00:00:24.480 --> 00:00:26.640
+Ali možnosti, ki jih prinaša generativna UI,
+
+12
+00:00:26.640 --> 00:00:28.600
+vplivajo na vse vrste vsebin?
+
+13
+00:00:28.600 --> 00:00:30.640
+Kako je z avtorstvom?
+
+14
+00:00:30.640 --> 00:00:32.760
+Je avtor podobe oseba,
+ki je izdelala računalniški program?
+
+15
+00:00:32.760 --> 00:00:34.440
+Ali pa je to program,
+ki je ustvaril podobe,
+
+16
+00:00:34.440 --> 00:00:35.680
+ki pred tem niso obstajale?
+
+17
+00:00:35.680 --> 00:00:37.560
+In – če nismo prepričani,
+ali je podoba resnična
+
+18
+00:00:37.560 --> 00:00:40.280
+– ali to pomeni,
+da fikcija ne obstaja več?
+