AI News: Brace Yourself for the Coming AI Storm!

Matt Wolfe
1 Mar 202423:46

Summary

TLDR2024年3月のAI界には興奮が巻き起こっています。昨年の大きなAI関連発表に続き、さらに多くの進展がありました。今週は、AIによるビデオ生成技術の大きな進歩が見られ、SoraやP Labsのアップデート、Runwayのプラットフォーム改良、Alibabaからの新しい研究などが発表されました。また、LTX StudioがAIを利用したビデオ生成プラットフォームを発表し、Google DeepMindやAdobeもAI技術で新たな取り組みを公開しました。Appleは自動運転車プロジェクトを終了し、AIへの注力を強化。MicrosoftはAI分野でのパートナーシップを拡大し、AI技術の社会への影響がさらに議論されています。

Takeaways

  • 😲 Many major AI announcements happened in March 2023, more big news expected in March 2024
  • 🎥 New video generation capabilities shown by Sora, Anthropic, RunwayML and others
  • �$$ Tyler Perry canceled plans for $800 million movie studio after seeing Sora AI video tech
  • 🕹 Google DeepMind AI can generate simple platformer video games from images and drawings
  • 🎵 Adobe showed AI that can generate music from text prompts like 'happy dance'
  • 😊 Idiogram 1.0 generates images with better realism and text handling
  • 🚘 Apple has stopped development on its car to focus more on AI
  • ☁️ Microsoft partners with French AI company Anthropic
  • 💼 Clara used AI chatbot to replace 700 human customer support jobs
  • 👀 Lenovo showed transparent laptop screen, but unclear on useful applications

Q & A

  • 2024年3月にはAI分野で何が期待されていますか?

    -2024年3月は、前年に引き続きAI分野での重大な発表や進展が期待されている月とされています。

  • Soraとは何ですか、どのような進展がありましたか?

    -SoraはAIビデオ生成ツールであり、マーク・ブラウンリーが自身でプロンプトしたビデオ例(犬が駐車場を歩くビデオなど)が公開され、大きな注目を集めています。

  • P Labsはどのようなアップデートを行いましたか?

    -P Labsはビデオ生成モデルをアップデートし、ビデオにリップシンクを追加する機能を導入しました。

  • Runwayはどのようなマイナーアップデートを行いましたか?

    -Runwayはユーザーインターフェースを改善し、動画内の特定エリアへのアニメーション効果をより細かく制御できる新機能を追加しました。

  • Alibaba Groupからの研究ではどのような技術が紹介されましたか?

    -Alibaba Groupの研究では、「Emo Emote Portrait Alive」という技術が紹介され、オーディオに合わせて表情が変化するリアルタイムのポートレートビデオを生成することができます。

  • LTX Studioとは何で、どのような機能を提供していますか?

    -LTX StudioはAIを活用してフルビデオを生成するプラットフォームで、シングルプロンプトから複数のシーンやショットを生成し、編集やカスタマイズが可能です。

  • Tyler Perryはなぜ映画スタジオの拡張計画を中止しましたか?

    -Tyler PerryはSoraの能力を目の当たりにして、映画産業の未来について再考し、約8億ドルの映画スタジオ拡張計画を中止しました。

  • Google DeepMindの「Genie」とは何ですか?

    -「Genie」はGoogle DeepMindによるプラットフォーマーゲームを生成するAIで、人々がプレイするビデオゲームの映像から学習し、新しいゲームを作成できます。

  • Adobeが発表した新しい音楽生成AIツールについて教えてください。

    -Adobeは「Project Music Gen AI Control」という新しい音楽生成AIツールを発表し、テキスト記述からオーディオを生成することが可能です。

  • AppleはAI分野でどのような動きを見せていますか?

    -Appleは自動運転車のプロジェクトを中止し、そのリソースを生成型AIへとシフトしており、Siriの大幅なアップデートや開発プラットフォームにおけるAI機能の拡充が期待されています。

Outlines

00:00

🌀 AIビデオ革新の進展

2024年3月のAI界では、前年に続き重大な発表があると予測されています。特にAIビデオ生成技術が大きく進歩し、SoraやP Labsからの新しいビデオ生成例や機能が公開されました。これらの技術は、リアルタイムでのリップシンクや感情を反映した顔の動きなど、ビデオ生成の新たな可能性を示しています。LTX StudioはAIを用いたビデオ制作プラットフォームを提供し、シングルプロンプトから複数のシーンを生成する能力を備えています。

05:03

🚀 AI技術の急速な進化

AIビデオ空間の進化は驚異的で、LTX Studioはユーザーに全面的なビデオ生成プラットフォームを提供しています。Tyler PerryはSoraの能力を目の当たりにして、巨額の映画スタジオ計画を中止するに至りました。Google DeepMindのGenieはAIによるプラットフォーマーゲーム生成を実現し、AdobeはAIを活用した音楽生成ツールを発表しました。また、idiogram 1.0ではリアリズムとテキストの正確性が大幅に向上しています。

10:04

🌐 AI画像生成技術の新展開

idiogram 1.0のリリースにより、AI画像生成技術のリアリズムとテキスト生成能力が向上しました。Googleは人物画像生成での誤りを認め、改善を約束しています。Appleは自動運転車プロジェクトを終了し、代わりに生成型AIに注力することを発表しました。MicrosoftはAIを活用した財務管理ツールを導入し、Tumblrとwordpress.comが訓練データ提供でAI企業と交渉していると報じられました。

15:06

🔍 AIと社会への影響

MicrosoftはAIを活用した財務管理ツール「Co-pilot for Finance」を発表しました。Tumblrとwordpress.comは訓練データの提供でAI企業と交渉中であり、Googleもstack Overflowと協力しています。また、open AIとNew York Timesの間で著作権侵害を巡る訴訟が進行中です。AIの進歩は、業務の効率化や新たな技術開発をもたらす一方で、雇用への影響やプライバシーの問題も提起しています。

20:09

🔬 透明なラップトップディスプレイとAIの未来

Lenovoから発表された透明なラップトップディスプレイは注目を集めていますが、その実用性には疑問が残ります。AI技術の進化は続いており、特にビデオ生成やAI画像生成技術における進歩が目覚ましいです。企業はAIを活用した製品やサービスの開発に積極的であり、社会や雇用への影響も深刻に考えられています。AIの未来は明るいものの、その影響を理解し適切に対応することが求められます。

Mindmap

Keywords

💡SORA

ソラはAIを利用したビデオ生成プラットフォームです。このビデオでは、ソラがより高度なビデオを生成できることが示され、多くの人々の関心を引いていることが説明されています。

💡LTXスタジオ

LTXスタジオは、AIを使ってビデオを生成するプラットフォームです。このビデオのスポンサーとして紹介されており、1つのプロンプトから複数のシーンやショットを生成できるなどの機能があると述べられています。

💡AIビデオ生成

この分野は非常に速く進歩しており、以前に比べてはるかに高度なビデオをAIで生成できるようになっていることがこのビデオで指摘されています。

💡Genie

Google DeepMindが開発したAIで、プラットフォーマー・ゲームを生成できることが示されています。

💡アップル

アップルが自動運転車の開発を中止し、代わりにAI分野に注力することが発表されたと報じられています。

💡Mistl AI

フランスのAI企業で、マイクロソフトがパートナーシップを結んだと発表されました。AI分野でのマイクロソフトの優位性をさらに高めることが期待されています。

💡大規模言語モデル

このビデオでは、OpenAIやMicrosoftなどの企業が自社のAIモデルを訓練するために大量のデータを取得しようとしていることが報告されています。

💡AIによる代替

Clara社がAIチャットボットを導入し、700人の従業員を解雇した例が取り上げられ、AIが人間の仕事を奪うリスクがあることが指摘されています。

💡透明ディスプレイ

レノボが透明なラップトップディスプレイを開発したが、その用途が不明確であるとして批評されています。

💡3月

2023年3月に重要なAI関連の発表が相次いだことから、2024年3月もAI分野での波乱含みの月になることが予想されていると述べられています。

Highlights

We got even more video examples from Sora and what it's capable of including some that Mark has brownley got to prompt himself.

P Labs also got an update to their video generation model this week allowing us to add lip syncing to our videos.

Runway also got some minor updates to their platform we've got a new UI here in Gen 2.

This isn't just a lip sync where you can upload an audio file and it will match the lips to the audio it actually understands the emotions behind the audio as well and causes the facial expressions in the face to move a little more closely with the emotion that comes with the audio.

Adobe unveiled Project Music gen AI Control a platform that can generate audio from text descriptions like happy dance or sad Jazz.

Deep mind's new AI makes games from scratch where he does an almost 8 minute breakdown of this model.

Idiogram just released idiogram 1.0, which is a huge improvement over the previous version of idiogram it's so much better at realism.

Google came out and a iiz they said Gemini image generation got it wrong we'll do better we've acknowledged the mistake, and temporarily paused image generation of people in Gemini while we work on an improved version.

The rumor is that we're going to get a big upgrade to Siri this year, where if you use an Apple device Siri might work a little bit more like what we expect from something like chat GPT.

Microsoft invested in open AI they invested in meta and meta's open- Source large language models they've partnered with hugging face and now they've partnered with mistol who has both closed and open AI models.

Microsoft this week also rolled out co-pilot for finance this is really a suite of AI tools inside of Microsoft 365 that helps the people that manage the finances for a company.

In my opinion it's definitely one of the downsides one of the sort of realizations that I struggle with around AI personally but at the same time I think everybody should be aware that this is happening everybody should be preparing for this.

The CEO of Clara claims that he's using AI to do the work of 700 people after he actually laid off 700 people back in 2022.

Open AI is asking the courts to dismiss this case because of that and the fact that normal people do not use open ai's products this way.

I talked about this one a few weeks ago but open AI is currently in a lawsuit with the New York Times.

Transcripts

play00:00

there seems to be a storm of breu in the

play00:01

AI world if you remember it was March of

play00:05

2023 that some of the biggest AI

play00:08

announcements that just totally broke

play00:10

everybody's brains happened in March of

play00:13

last year this video is coming out on

play00:15

the first day of March of 2024 and I

play00:17

think we're in for another wild ride

play00:19

this month so let's get into all of the

play00:22

interesting and exciting things that

play00:23

happened in the world of AI this week

play00:26

this video is sponsored by LTX Studio A

play00:29

really awesome new AI generation

play00:32

platform and I'll tell you more about

play00:34

that in just a few minutes here but this

play00:36

week saw some huge progress in the world

play00:38

of AI video I don't really want to dive

play00:40

too deep into all of the announcements

play00:42

that came out in the world of AI video

play00:44

because I did make another video this

play00:45

week you can find it on my channel it

play00:47

was called I was wrong about AI video

play00:50

that video specifically breaks down all

play00:53

of the really exciting AI video

play00:55

generation announcements that we had

play00:57

this week but for those of you that just

play00:58

want the tippy top of the iceberg and a

play01:01

simple recap of what we saw this week we

play01:04

got even more video examples from Sora

play01:07

and what it's capable of including some

play01:09

that Mark has brownley got to prompt

play01:11

himself like this video of a dog walking

play01:14

in a parking lot and a 3D printer making

play01:16

something here and a product reviewer

play01:18

surrounded by gadgets and of course Sora

play01:21

is still managing to blow people's minds

play01:24

but for the most part people just want

play01:25

to get their hands on it like me I want

play01:27

to play with it P Labs also got an

play01:29

update to their video generation model

play01:31

this week allowing us to add lip syncing

play01:34

to our videos you can either do a text

play01:36

to speech model type in what you want

play01:38

your character to say or upload your own

play01:40

audio file and it will speak out what

play01:43

came from the audio file wait hold up

play01:45

hold up hold up these things can talk

play01:48

for real Runway also got some minor

play01:51

updates to their platform we've got a

play01:53

new UI here in Gen 2 and if we click on

play01:56

the motion brush here to sort of get

play01:59

more control over over how we want to

play02:00

animate this image there's this new

play02:02

autod detect area so we can actually let

play02:05

the AI detect what areas we want to

play02:08

affect in the video so if I wanted to

play02:10

just animate my face before I could sort

play02:13

of draw over my face and mask it out now

play02:16

it sort of finds my face for me so I can

play02:18

just click that tell to animate just my

play02:20

face and then watch as it messes with

play02:23

just that portion of the image we got

play02:26

this research out of Alibaba group

play02:28

another thing that we don't have access

play02:29

too but looks really cool and I'm

play02:31

itching to get my hands on called emo

play02:34

emote portrait alive generating

play02:35

expressive portrait videos with audio to

play02:37

video diffusion model under weak

play02:39

conditions this isn't just a lip sync

play02:41

where you can upload an audio file and

play02:44

it will match the lips to the audio it

play02:46

actually understands the emotions behind

play02:48

the audio as well and causes the facial

play02:51

expressions in the face to move a little

play02:53

more closely with the emotion that comes

play02:55

with the audio here's an example of

play02:57

miror Mora from open AI speaking and

play03:00

this image actually using those same

play03:03

vocals so check this out and uh maybe we

play03:06

have several of them and maybe they all

play03:08

have different competences here's

play03:10

another example where this image on the

play03:11

left was uploaded and then a music

play03:14

sample was provided and then on the

play03:15

right we can see the animation that was

play03:18

[Music]

play03:25

generated and it's just way more than

play03:27

lips syncing it's eye blinks it's the

play03:30

head movement it really sort of follows

play03:32

along to the emotion of the audio that

play03:35

it was given and there's tons of

play03:37

examples I can't really play them all

play03:39

because a lot of it is copyrighted music

play03:40

and I will get in trouble but I will

play03:42

link up this research in the description

play03:44

and once again I did do a deep dive

play03:47

video on all of the latest AI video

play03:49

announcements that have come out

play03:50

recently called I was wrong about AI

play03:52

video so check that one out if you

play03:54

haven't already we also got news this

play03:56

week out of L Trix who happens to be the

play03:58

sponsor of this video video If you're

play04:00

not familiar with lyrix they're the

play04:01

company behind really cool tools like

play04:04

motion leap video leap and photo leap

play04:06

well now they've got LTX Studio a

play04:09

platform that lets us generate full

play04:10

videos with AI I'm talking about you

play04:12

give it a single prompt and it generates

play04:14

multiple scenes with multiple shots

play04:17

within that scene consistent characters

play04:19

controllable lighting music sound

play04:22

effects everything that goes into

play04:24

creating a video you can generate with

play04:26

one prompt and once you have all of your

play04:27

scenes and all of your shots you can

play04:29

edit any piece you want on a shotby shot

play04:32

basis you have complete control so if

play04:34

you want to make a green car red simply

play04:36

prompt it and your green car is now red

play04:38

need to rearrange shots within your

play04:39

storyboard you'll be able to quickly

play04:41

rearrange any clip and put it wherever

play04:44

it needs to be within your video this is

play04:45

a tool that Prides itself on consistent

play04:48

characters consistent lighting and can

play04:50

even do text and titles so if you have

play04:52

an idea for a story and you want to

play04:54

bring that story to life with AI video

play04:56

LTX Studio can help you do that with a

play04:59

complete a toz video creation platform

play05:02

this is unlike anything we've seen

play05:04

before and really really shows off how

play05:07

quickly this AI video space is moving

play05:10

what we're able to do these days is

play05:12

pretty dang mindblowing and LTX studio

play05:15

is making it easy all in one single

play05:18

platform right now you can get on the

play05:20

wait list to join LTX Studio when it's

play05:22

available and I have a special link in

play05:24

the description that'll even give you

play05:26

early access so make sure you use the

play05:28

link in the description for this video

play05:31

and as soon as it's ready you'll be one

play05:32

of the early ones to play with LTX

play05:34

Studio thank you so much to LX and their

play05:37

new LTX Studio platform for sponsoring

play05:39

this video we've got so many more cool

play05:41

things to share around LTX studio and

play05:44

I'll be talking about them a lot in

play05:45

future videos because this is a platform

play05:47

that I'm really really excited about and

play05:50

jumping back to the topic of Sora for

play05:52

just a second here this was an

play05:53

interesting news story where Tyler Perry

play05:57

saw Sora and then decided to to cancel

play06:00

the creation of an $800 million movie

play06:03

studio that he was working on so he was

play06:06

ready to invest almost a billion dollars

play06:09

into expanding his Studio saw what Sor

play06:12

was capable of and went I don't know if

play06:14

this is still the future of the film

play06:16

industry let's put a pause on this this

play06:19

week also saw some research come out of

play06:20

Google Deep Mind called Genie which is

play06:23

an AI that can generate sort of

play06:26

platformer games here's some of the

play06:28

examples on the page where you can see

play06:30

these little sort of NES era looking

play06:34

platformer games and these were all

play06:36

created by giving the AI a bunch of

play06:39

video footage of people playing through

play06:41

games and then it figured out how to

play06:43

recreate these games without any

play06:46

additional extra input they were able to

play06:47

start with reference images like this

play06:49

image here and generate a movable

play06:52

character in this world they were even

play06:55

able to upload you know children's art

play06:57

and cutouts and turn those into little

play07:00

playable games that you can play as well

play07:02

heck they even did it with real world

play07:03

images now you can tell these are very

play07:05

small resolution and they're only

play07:06

running at one frame per second so you

play07:09

probably wouldn't want to play these

play07:11

games yet but the fact that it can learn

play07:13

from just a sample of other video games

play07:16

that are out there create a new game

play07:19

from an image or from a drawing and then

play07:22

actually put a movable character in that

play07:25

world pretty impressive it's the worst

play07:27

it's going to get right now and if you

play07:28

saw how fast things like Dolly got to

play07:31

where they are with Dolly 3 now and how

play07:33

quickly we went from things like model

play07:35

scope to where we're now seeing Sora

play07:38

this is just that early moment for these

play07:41

little platform video games where who

play07:44

knows a year from now some of these

play07:46

could look like they were generated with

play07:47

Unreal Engine 5 who knows if you asked

play07:49

me a few months ago I would say that was

play07:51

a huge stretch but at the pace that

play07:53

things are advancing and the exponential

play07:55

curve were on I don't think it's out of

play07:56

the question anymore if you want a deep

play07:58

dive and see some more examples of what

play08:00

this Genie is capable of of course I'm

play08:02

going to link the research up below so

play08:03

you can check out some of these videos

play08:05

but hold on to your papers my friend

play08:07

because one of my favorite channels

play08:08

two-minute papers just did a breakdown

play08:10

video called Deep mind's new AI makes

play08:12

games from scratch where he does an

play08:14

almost 8 minute breakdown of this model

play08:17

so you can see more of it I'll likely do

play08:18

a deeper dive myself in a future video

play08:21

but if you want to learn more that's a

play08:23

great start what a time to be alive and

play08:25

since we've been talking about

play08:26

multimedia so far in this video Let's

play08:29

Talk about Adobe and how they just

play08:30

revealed a new generative AI tool for

play08:33

music during the hot pod Summit in

play08:35

Brooklyn Adobe unveiled Project Music

play08:38

gen AI Control a platform that can

play08:40

generate audio from text descriptions

play08:42

like happy dance or sad Jazz here's a

play08:45

little sneak peek of what it can do I

play08:47

can uh take an input Melody it sounds

play08:49

like

play08:49

[Music]

play08:52

this and then you can generate the

play08:54

accompaniment in this case some some

play08:56

film

play08:56

[Music]

play08:58

music

play09:00

that's cool that's just from a prompt

play09:01

yeah yeah do you want to hear something

play09:03

else yeah let's let's do hip-hop all

play09:05

right one

play09:07

[Music]

play09:09

sec okay check out this other example

play09:12

where I can control the intensity up or

play09:16

down oh yeah no that was intense the

play09:19

article here says that project music gen

play09:22

control may be made publicly available

play09:24

at some future date but for now it's

play09:26

firmly in the research stage but

play09:27

something like this would be so cool for

play09:29

background music on things like YouTube

play09:31

videos podcasts or even your generative

play09:35

games that you might be making I love

play09:37

seeing all of this cool AI research

play09:38

that's in the works but I'm always

play09:40

bummed out that we can't get our hands

play09:42

on it and play with it ourselves yet CU

play09:44

that's what we all really want let's

play09:46

jump over to AI art for a second because

play09:48

idiogram just released idiogram 1.0

play09:51

which is a huge improvement over the

play09:54

previous version of idiogram it's so

play09:56

much better at realism it's so much

play09:58

better at at adding text to your images

play10:01

I got a little bit early access I've

play10:02

been playing with this one for a couple

play10:04

weeks now but it does some really good

play10:06

work by log into my account here you can

play10:08

see some of the stuff I generated with

play10:10

1.0 I generated some images that say

play10:12

subscribe to Matt wolf on YouTube you

play10:15

can see all of the spelling is right and

play10:16

it actually generated four images you

play10:18

can see my prompt was a group of wolves

play10:20

holding up a sign that says subscribe to

play10:22

Matt wolf on YouTube here's another one

play10:24

that it made nailed the text another one

play10:26

once again nailed the text and the

play10:29

fourth one so all four of the iterations

play10:31

that it generated nailed the text I did

play10:34

ask it to generate the word Mr eow

play10:37

spelled out in the Cloud's conceptual

play10:39

art 3d render and well it missed the F

play10:42

it says Mr ELO and if you look at all of

play10:44

the generations all of them made the

play10:46

same mistake it got the concept right

play10:49

just spelled the word wrong but if we

play10:51

look at some of the other images that

play10:52

people have generated here we can see

play10:54

that it could generate famous people and

play10:57

this cartoony style it does really cool

play11:00

high contrast colors similar to what mid

play11:02

journe is capable of and it's even

play11:04

gotten so much better at realism

play11:07

videogram did add some paid plans but

play11:09

the free plan still allows you to

play11:12

generate up to 100 images a day and

play11:14

since we're on the topic of AI image

play11:16

generation if you remember last week I

play11:18

showed you that Google ran into some

play11:20

issues where people would ask for like

play11:22

historical images and it would get like

play11:24

the race wrong of the image well Google

play11:27

came out and a iiz they said Gemini

play11:30

image generation got it wrong we'll do

play11:32

better we've acknowledged the mistake

play11:33

and temporarily paused image generation

play11:35

of people in Gemini while we work on an

play11:38

improved version so they realize that

play11:41

when you're asking for historical

play11:42

figures they should be the actual

play11:44

ethnicity of the historical figure that

play11:46

you asked for now when we can expect

play11:48

these changes to be made Google claims

play11:50

they should be fixed within weeks so I

play11:52

would imagine within a week or two of

play11:54

this video coming out we'll get Gemini

play11:56

images again and we'll go back through

play11:58

and test it and see if we can actually

play12:01

get it to generate what we want it to

play12:03

generate we're starting to hear more and

play12:04

more AI news Bubble Up out of apple as

play12:07

well since 2017 we've been hearing that

play12:10

Apple was building the Apple car rumor

play12:12

had it that it was going to be a

play12:13

competitor to like a Tesla however this

play12:15

week Apple finally announced that

play12:18

project is dead they're moving on past

play12:19

the apple car they're not going to

play12:21

continue to work on it how is that

play12:23

relevant to AI according to Bloomberg

play12:26

most of the team's workers that were

play12:28

working on that car are moving over to

play12:30

generative AI initiatives it sounds like

play12:33

there might be some layoffs at Apple

play12:35

over this but for the most part they're

play12:37

just sort of reiting priorities and

play12:39

instead of focusing on these autonomous

play12:42

vehicles they're putting the focus into

play12:44

generative AI instead and Tim Cook even

play12:47

further confirmed that by saying that

play12:49

Apple will break new ground in

play12:51

generative AI this year Apple had their

play12:53

annual shareholder meeting this week and

play12:56

during the company's q1 earnings call

play12:59

cook said Apple was working internally

play13:00

with Gen AI but it was taking a slower

play13:03

more deliberate approach to customer

play13:05

facing incarnations of the technology

play13:08

indeed Apple's only briefly mentioned

play13:10

generative AI in its recent press

play13:12

conferences and announcements such as

play13:14

when it introduced new autocorrect and

play13:16

text prediction features in iOS last

play13:19

fall the rumor is that we're going to

play13:21

get a big upgrade to Siri this year

play13:24

where if you use an Apple device Siri

play13:27

might work a little bit more like what

play13:29

we expect from something like chat GPT

play13:31

apple is also said to be exploring AI

play13:33

powered features to allow users to

play13:35

automatically generate presentation

play13:36

slides and Keynotes and playlists in

play13:38

apple music as well as generative AI

play13:40

powered coding suggestions in xcode the

play13:43

company's app development platform we

play13:45

also got the news this week that

play13:46

Microsoft was partnering with mistol AI

play13:49

the French AI company that has built

play13:51

some of the best open-source AI models

play13:54

we have available today mistl AI will

play13:57

have access to a z years's Cutting Edge

play13:59

AI infrastructure and Cloud computers

play14:01

while further boosting Microsoft's

play14:04

absolute dominance in the AI space right

play14:06

now if you're keeping score Microsoft

play14:08

invested in open AI they invested in

play14:11

meta and meta's open- Source large

play14:13

language models they've partnered with

play14:15

hugging face and now they've partnered

play14:18

with mistol who has both closed and open

play14:21

AI models so it just seems like

play14:23

Microsoft is sort of spreading the deck

play14:25

diversifying their risk they're working

play14:27

with open Ai and their closed model

play14:29

they're working with meta and their open

play14:31

source models they're working with mistl

play14:33

who has both open and closed models and

play14:36

they're working with hugging face which

play14:38

is becoming sort of like the central hub

play14:41

for where all of this AI development is

play14:43

happening Microsoft really has their

play14:45

toes in every single pond so no matter

play14:49

which sort of scenario plays out in the

play14:52

AI World whether these large closed

play14:54

models win or open source wins Microsoft

play14:57

wins regardless because they're all

play14:59

using Microsoft's compute and Microsoft

play15:02

hedged their bets and sort of bet on all

play15:05

potential outcomes and since we're on

play15:07

the topic of Microsoft Microsoft this

play15:09

week also rolled out co-pilot for

play15:11

finance this is really a suite of AI

play15:14

tools inside of Microsoft 365 that helps

play15:17

the people that manage the finances for

play15:19

a company so it helps Financial analysts

play15:22

quickly conduct a variance analysis in

play15:24

Excel using natural language prompts it

play15:27

simplifies the reconciliation process in

play15:29

Excel provides a complete summary of

play15:31

relevant customer account details and

play15:33

Outlook enables customers to turn raw

play15:35

data in Excel into presentation ready

play15:38

visuals and reports if you're somebody

play15:39

that sort of manages the financial side

play15:41

of a business this is probably something

play15:43

you're going to want to look into a lot

play15:45

of these large language models are

play15:46

getting new resources to train their

play15:49

models on as well this week it came out

play15:51

that tumblr's owner is striking deals

play15:53

with both open Ai and mid Journey for

play15:56

training data owner of Tumblr and

play15:58

wordpress.com is in talks with AI

play16:00

companies mid journey and open AI to

play16:02

provide training data scraped from users

play16:04

posts at the moment this is sort of

play16:06

still rumor this is based on an

play16:09

anonymous source and neither open AI or

play16:12

mid Journey have actually confirmed this

play16:14

but there is a lot of content on

play16:16

wordpress.com and Tumblr sites that

play16:19

these platforms can potentially train on

play16:21

although let's be honest wordpress.com

play16:24

and Tumblr have become a lot less

play16:26

relevant in recent years which is also

play16:28

probably part of the reason why Tumblr

play16:31

and wordpress.com want to sort of

play16:33

monetize this data because it's a new

play16:36

play a new rabbit hole that they can go

play16:38

down by allowing their data to be

play16:41

trained into AI platforms so they can

play16:43

make more money because their platforms

play16:45

seem to slowly be going downhill and I'm

play16:48

talking about wordpress.com not you know

play16:50

self-hosted wordpress.org blogs that's a

play16:53

different story don't think those are

play16:54

going downhill the hosted blogs you just

play16:56

don't hear about nearly as much as you

play16:58

used to and then there's Google they

play16:59

seem to be working with stack Overflow

play17:01

to help provide Gemini cloud with

play17:04

training data as you might already know

play17:06

Microsoft owns GitHub so Microsoft has

play17:10

all of the data that's inside of GitHub

play17:12

that it can use to train the various

play17:14

models that Microsoft has available to

play17:17

them at one point stack Overflow

play17:19

launched overflow AI but I don't think

play17:21

it ever really gained any traction not

play17:23

many people talked about it it wasn't

play17:25

really that big of a hit and stack

play17:27

Overflow Al so let's be honest kind of

play17:29

feels a little bit like a sinking ship

play17:31

people don't really need to go to stack

play17:33

Overflow as much to try to debug coding

play17:35

problems as they used to because now

play17:37

they have chat Bots they've got GitHub

play17:39

they've got GitHub co-pilot they've got

play17:41

Amazon code whisper they have all these

play17:43

other ways to debug that people just go

play17:45

to stack Overflow a little bit less than

play17:47

they used to stack Overflow needs to

play17:49

figure out a way to stay relevant and if

play17:51

people aren't going to go to their site

play17:53

well maybe they can sell the data that's

play17:55

on the site to someone like Google and

play17:58

actually continue to monetize what

play18:00

they've built in this new era of AI that

play18:03

we're entering into I talked about this

play18:05

one a few weeks ago but open AI is

play18:07

currently in a lawsuit with the New York

play18:09

Times the New York Times is claiming

play18:11

that open AI regurgitated one of its

play18:13

articles of verbatim open AI claimed

play18:16

well you had to really really sort of

play18:18

dial in the prompt and even give it some

play18:21

copied and pasted sentences from that

play18:23

original article to get it to prompt

play18:25

your article verbatim it says here the

play18:28

New York times used deceptive prompts to

play18:30

get chat GPT to regurgitate its content

play18:32

so open AI is asking the courts to

play18:34

dismiss this case because of that and

play18:37

the fact that normal people do not use

play18:39

open ai's products this way this week

play18:41

the CEO of Clara claims that he's using

play18:44

AI to do the work of 700 people after he

play18:48

actually laid off 700 people back in

play18:51

2022 the CEO explained that in the chat

play18:54

bot's first four weeks it had handled

play18:55

about 66 of its customer support or

play18:59

about 2.3 million chats and it even

play19:02

scored an equivalent in customer

play19:04

satisfaction when compared to humans

play19:06

even exceeding humans in some cases the

play19:09

CEO said so while we're happy about the

play19:11

results for our customers our employees

play19:13

who have developed it and our

play19:15

shareholders it raises the topic of the

play19:17

implications it will have for society

play19:20

and this is one of those cases where an

play19:22

AI legitimately has replaced about 700

play19:26

jobs and we're going to hear more and

play19:28

more stories like this in the coming

play19:30

months and years in my opinion it's

play19:32

definitely one of the downsides one of

play19:35

the sort of realizations that I struggle

play19:37

with around AI personally but at the

play19:40

same time I think everybody should be

play19:42

aware that this is happening everybody

play19:44

should be preparing for this at the end

play19:46

of the day you're so much better off

play19:48

knowing what's coming knowing what tech

play19:50

exists and being looped in than getting

play19:53

completely blindsided when it does knock

play19:55

you out it's a sad fact that a lot of

play19:57

people will lose their jobs because AI

play20:00

can replace them but I'm really really

play20:02

hoping that us as humans as Humanity

play20:05

find new ways to provide value to other

play20:08

people and finally this video has been

play20:11

circulating on X about this transparent

play20:14

laptop display from Lenovo and I find

play20:18

this really interesting I think the tech

play20:20

is really really cool like it's awesome

play20:22

to see this laptop and a transparent

play20:25

display through the laptop it's very

play20:27

future istic very cool looking however I

play20:31

have no clue what the use case for this

play20:33

is like why would you want a transparent

play20:35

laptop display we already have

play20:37

transparent monitors we saw them at CES

play20:40

this year they're great for like

play20:41

storefront windows that you can turn

play20:43

into monitors they're great for display

play20:46

cases and things like that but what's

play20:47

the use case for a transparent laptop

play20:49

monitor to me it looks like it would

play20:51

just be harder to read you lose all

play20:53

privacy cuz if somebody's sitting across

play20:54

the table from you they could just see

play20:56

right through it and I don't I don't

play20:58

know it just feels gimmicky I love it it

play21:00

looks cool but I don't really want one

play21:02

cuz I don't know what I would use it for

play21:04

or what the actual real use case of it

play21:07

is I thought I would show it off cuz I

play21:08

thought it was interesting and well it

play21:10

looks cool but maybe you have an idea of

play21:12

what that could be used for let me know

play21:13

in the comments if you know what a

play21:15

transparent monitor on a laptop would

play21:17

really be used for because I'm really

play21:19

struggling to come up with the best use

play21:21

case and that's it that's what I got for

play21:23

you this week it's been another really

play21:25

really big week in AI the previous two

play21:27

weeks before this where a little bit

play21:28

bigger a little bit crazier news came

play21:31

out of it this is another one of those

play21:32

sort of marginal weeks where we got some

play21:34

little updates here and there but again

play21:36

this video is dropping on March 1st and

play21:38

last year March was a whirlwind so we're

play21:41

about to head into some exciting times

play21:44

the storm is just getting started I'm so

play21:46

excited I'm here for it I'm going to be

play21:48

reporting on it telling you about it

play21:50

keeping you looped in and I'm really

play21:51

excited to see what rabbit holes this

play21:53

takes us down one quick announcement I

play21:55

am going to be at GTC in San Jose that's

play21:58

nvidia's event they have both an

play22:00

in-person event and a virtual version

play22:02

cost money to go to the inperson event

play22:04

the virtual version you can access for

play22:06

free and if you register you're entered

play22:08

to win an RTX 480 super GPU from Nvidia

play22:13

I've actually got it here sitting next

play22:15

to my desk right now still in the box if

play22:17

you win I'm going to send this exact

play22:19

unit to you all you have to do is

play22:21

register for the virtual event using my

play22:24

link in the description I will get the

play22:26

names of the people that registered from

play22:28

Nvidia and we will pick a winner around

play22:30

the time of the event so if you're not

play22:32

registered for it it's totally free you

play22:35

could win an Nvidia 4080 super just by

play22:37

registering for it I will make sure

play22:39

again that link is in the description

play22:41

also if you haven't already check out

play22:42

future tools. where I curate all of the

play22:45

coolest AI tools that I come across I

play22:47

sort them for you so you can drill down

play22:49

to just the tools that you're most

play22:51

interested in I keep the AI news page up

play22:54

to date on pretty much a daily basis and

play22:56

I have a free a AI newsletter or if you

play22:59

sign up for the newsletter I'll also

play23:01

hook you up with the AI income database

play23:03

a database of really cool ways to make

play23:05

money using AI it's totally free you

play23:08

just got to join the free newsletter

play23:10

over at futur tools. and I will hook you

play23:13

up and if you like this video you want

play23:14

to stay in the loop with the latest AI

play23:16

News latest AI tutorials research and

play23:18

just nerd out with me about really cool

play23:21

Ai and Tech stuff make sure you like

play23:23

this video And subscribe to this YouTube

play23:25

channel and I will make sure more videos

play23:27

like this show up in your YouTube feed

play23:29

thank you so much for tuning in today I

play23:31

really really appreciate you thank you

play23:33

so much to LTX studio and lightricks for

play23:35

sponsoring this video that is a really

play23:37

cool platform that I am super excited

play23:39

about check that out get on the wait

play23:41

list if you haven't already and thanks

play23:43

once again for tuning in I'll see you in

play23:45

the next video bye-bye