AI News: Get Ready, The World is About to Change
HTML-код
- Опубликовано: 8 июн 2024
- Here's the AI news from the past week you might have missed. Check out Tubebuddy here: bit.ly/48q07ey
Discover More From Me:
🛠️ Explore hundreds of AI Tools: futuretools.io/
📰 Weekly Newsletter: www.futuretools.io/newsletter
😊 Discord Community: futuretools.io/discord
🐤 Follow me on Twitter: / mreflow
🧵 Follow me on Instagram: / mr.eflow
🐺 My personal blog: mattwolfe.com/ (Currently Being Overhauled)
Resources From Today's Video:
Nvidia GTC: nvda.ws/3tLZAVR
Marcelo at SXSW: / 1767696317843677398
OpenAI Update: • Figure Status Update -...
GPT-4.5 Turbo Leak: the-decoder.com/openais-gpt-4...
Leadership Continuity: openai.com/blog/review-comple...
New Board Members: openai.com/blog/openai-announ...
AI Engineer Devin: news.futuretools.io/meet-devi...
AI Explained on Devin: • AI Agents Take the Whe...
Volo on Devin: • Devin AI Agent is WAYY...
Karpathy on Software: / 1767598414945292695
Olin on AI Space: / 1767651974823006584
Musk's xAI: news.futuretools.io/elon-musk...
Claude 3 Haiku: www.anthropic.com/news/claude...
SIMA in 3D: deepmind.google/discover/blog...
Character Reference: news.futuretools.io/midjourne...
Midjourney Ban: arstechnica.com/information-t...
Sound Effects Tech: news.futuretools.io/pika-labs...
Sponsorship/Media Inquiries: tally.so/r/nrBVlp
Mailing Address: 3755 Avocado Blvd Unit 287, La Mesa, CA 91941
#AINews #AITools #ArtificialIntelligence
Time Stamps:
0:00 Intro
0:03 AI Hate at SXSW
1:58 Figure 1 Robot
5:00 GPT 4.5 Potential Leak
6:51 The End of The Sam Altman Saga
9:02 Split-Testing on RUclips
11:03 Devin Autonomous AI Coder
15:44 WebGPT Autonomous Coder
18:33 Elon Musk to Open-Source Grok
19:17 Claude 3 Haiku
21:05 SIMA - AI That Plays Videogames
23:05 MidJourney Consistent Characters
24:46 Leonardo adds "Collections"
25:59 Nvidia GTC Conference
26:55 Final Thoughts 
Наука
A robot handing a man an apple that's an interesting aesthetic
Plus the robot voice mimic Steve Jobs.
Do you think the robot should have a female voice, and perhaps be called Eve?
Praise Skynet!!!
@@harrybarrow6222 I thought the same 😂Would have been funny to me, but people would probably take it far too serious.
@@Enu_Vibe no, not in tjat way although Jobs may have used the Adam n Eve referense for similar reasons
People booing are from the film and music industry !
yeah it's just a small film festival in Austin. Not the typical audience for AI
True, but why? If they have a good film or song idea then wouldn't AI help them make it bigger, faster, stronger?
F**k A.I. That's why.
@@dattajackbecause they’re not the brightest or logical. It’s just emotional and reactionary, and I can understand it.
No, it's people who can think past the current buzz and see the potential in the very near future.
"Stop resisting" is maybe not the best phrase to have in your presentation about AI in a post Star Trek TNG world.
Well, I do see parallels between AI-enthusiasts and the Borg.
Yeah, that lady must never be in front of people again
When you're privileged, ivy league, rich and out of touch you will behave that way.
Yeah, totally get it, and I agree with the sentiment, but we need more tact when talking to the doomers and luddites.
@@mc9723 Agreed. While it was admittedly a bit abrasive how they phrased it in their presentation, I think the sentiment is true. AI is here whether we like it or not, so we should try to embrace it as best as we can. But that doesn’t mean we can’t try to influence the direction it goes in or how it plays into our lives, which is certainly important when we have these big corporations being the ones leading the push.
3:30 that small adjustment of the drying rack as he finished is terrifying. That's not scripted..
I was actually waiting for it to happen. I believe it works this way (oversimplifying oc) he took a picture and when asked to put the objects where they belong he generated one or many pictures depicting this ordered scene so he just had to match reality with one of those images. As the vase is not properly in place it doesn't match to a 100% the image so the robot makes the adjustments that are in his "power" to make so reality matches more closely the reference image. (Powered by DALL•E I suppose)
Ya that is what stuck out to me when I saw this yesterday. As well as the way he passes the trash basket back to the guy. It’s the little stuff like that which is fascinating to see.
It's a closed-loop machine-learning system. The LLM selects the weights according to the task it must perform. (This was explained in a Reddit post from Figure 1.)
10:20 I don’t know why, but I find the idea of AB testing thumbnails to be so corny.
I mean, I guess if this is what you do for a living then whatever but maybe keep it to yourself.
I like to at least think that the people I’m watching, are just focused on making a good video and tossing it on RUclips and not more concerned abouta green thumbnail or whatever
@@jonnyguitar747you are onto absolutely nothing my guy just dont even comment bro. What we're all talking about is stuff thats too much for your little monkey brain to even comprehend pal
@Matt: Thanks for calling out grossly exaggerated headlines on RUclips (=AGI) and elsewhere, it's becoming a real problem and a bit of a nuisance. As consumers, we would like to see video titles that honestly reflect the actual content and not click bait in the distorted style of the "National Enquirer".
Great video btw Matt (as always)!
So you don't think "the world is about to change" is sensational then? The world will change through this tech, but we have at least 1-2 years to see this stuff mature. Even longer until industry integrated in a meaningful way. It's been a year since low disruption ChatGPT and I don't know anyone who uses it yet. There is zero bar for entry and folks still don't see the value.
AI art is being actively eschewed because it's easily identifiable and numb.
Don't get me wrong, the world WILL change, and rapidly so. But it's not on the brink today.
I'll see you in a month where RUclipsrs are still saying "omg the world is about to change", or "Holy hell, you won't believe what happened, AGI at our door".
This is SHOCKING!!!
@@francoisbertrand7612 ... especially considering that I'm an AGI bot 😂
@@bloopbleepnothinghere This evolution of AI is millions and millions of times faster than biological evolution could have ever been. The Industrial age was only 200 years ago. Cell phones, computers, and the internet haven't even existed a little over 50 years and nowadays they are everything and everywhere. AI has taken off not even 10 years ago and now and it's also everywhere and its just getting faster and faster. There will soon be the day when AGI is no longer science fiction and changes the world and our fate in ways we can't fathom
@@Prodbyhope.mp3 all takes human amount of time to ingest and apply these technologies though. AI is moving at warp speed but tomorrow Hollywood will still kick out another blockbuster the same way they did yesterday. We’ve got years until Hollywood is fully integrated with production ready AI tools. Same goes for software. None of this stuff is end to end business ready yet. First adopters will be disruption, but displacement takes longer. The internet did not change the recording industry over night. The world is changing, but the revolution will be human paced.
The luddites were shaking their flint axes while ululating wildly.
Banging on the monolith
That first robot definitely needs some Halls
1:30 At least the video didn't say "Resistance is futile!"
That's about the 28493nd video with a title like "the world is going to change!" in a week for me. 😄
This reply to your comment is about to change the world forever.
Its almost like...the world is gonna change
Interesting that the reply to “Can I have something to eat?” Was not just “Yes”.
The robot inferred that it was a request for food, and that the robot should supply it.
Agreed - this is so subtle. It would be great to have the system explain it's reasoning on this. I mean, of course you can have something to eat. But, as you mention, inferring it to mean it was a request of the robot itself, rather than a generic question.
Maybe it was that his hand was on the table? Maybe because the apple was there? Curious what it would say/do if there was no apple.
And that was just GPT4.
What about the concept of ownership? While the robot could see the apple, it had no idea of who the apple belonged to. This should have been clarified first rather than the robot assuming it had the right to give the apple away.
@@alexdeburie5261As the LMM outputs task plans, that means it's either instructed to do so in the prompt (system or otherwise), or the model is a fine-tuned version for the bot rather than the generic public one. Either way, with it having the context that tells it, basically, "here's how you'd perform an action: ", it makes sense that if it sees food and someone asks it "can I have something to eat", it means "can you give me something to eat".
Any time I see AI brought up in casual settings, people hate it. I can't wait for people to be as excited about AI as I am...
I feel you. Too many people don't understand that "training" is not "copying", and would rather be outraged than learn how it really works.
@@IceMetalPunk Problem is that too Many Artist see AI training as Copying Sadly.
The worst part is them using the ignorance of Lawmakers to harm AI development
I think the main issue is that people fear change and are not willing to put in the effort to adapt.
@BobLadif-ny8niI mean... that one is pretty easy. We've had robots in manufacturing for longer than any other industry, and AI is used in those robots fairly commonly. Which part of manufacturing do you think embodied AIs would struggle with?
People faced with losing their jobs, will try to fight in the only way they can.
too late
Im excited about nobody working. Maybe we can all finally relax and enjoy eachothers time here.
i'm more than happy to enjoy my time with you big boy@@investigator2016
🧨🔥💥🔫
@@investigator2016 in the current system, everything that you need to survive is owned by somebody else (food, clothing, shelter, land) and the only means for you to get by is through labor in exchange for money. so you are only useful in the accumulation of capital. The state depends through taxes on you being employable aswell. It's going to be interesting so see how the state reacts once the capital owning class has rendered more and more workers obsolete to increase profits.
The Figure 1 video is interesting but I wonder how the robot responds to the person who later returns and asks: hey! Where’s the apple I put there earlier?
The quality of the robot's answer to this question will depend directly on the RAG implementation they use. If the implementation is correct, the answer will be correct because the chat GPT api. will receive the context information that the robot already gave the apple.
It keeps the whole conversation in context, which is how it can answer things like "explain why you did that" without re-specifying what "that" means. So when he comes back and asks where's the apple, it'll know and say it gave it to him earlier.
Now, if by "later" you mean "after the conversation has been wiped", it may not know; unless they save conversations vector-wise for RAG purposes, in which case it may be able to pull that answer back into the new context from the logs.
@@IceMetalPunk for a price :)
@@IceMetalPunk For an agent to maintain the entire conversation is something completely wrong, since in a matter of hours of interaction any conversation could easily exceed the context window, which is why it is essential to store the conversation memories within a vector database, since at the same time Doing this allows you to be able to manage memory for multiple days, months or years and always continue to have a good context, which is why the implementation of RAG is important. Furthermore, if the agent is well designed, it should not use a context window that traps giant conversations, since when working with a giant context window this significantly increases costs in terms of computing time and consequently economic costs, in addition to The RAG is the system that is normally used to manage the retriving of the agents' memories.
@@oscarbertel1449 You're right, I shouldn't have said "the entire conversation", as that's not accurate. But with that said... the new paper from Google about "infini-attention", if I understand it right, offers a memory solution that's more robust than RAG in terms of being able to incorporate the entire conversation (as opposed to just re-looking-up bits it thinks might be relevant each inference), without an increase in inference cost. So that may be the way forward.
As an experienced coder myself, I am very impressed that Devin practices RTFM.😉
I found it interesting that ‘Figure’ said that it could see “cups and a plate”. Not “A cup and plates”.
I like how one comment actually counted all of the comments that said first and types "5'th"
83rd winner winner chicken dinner.
You don't know that Figure 01 was powered with GPT-4. I doubt they used the version we have access to today.
I was assuming the same. Or did they say somewhere which GPT version is used? I'd like to know... Maybe it is even a special version for Figure 1, I mean, not just simply modified... (I am too tired, don't know if this expression fits right now as English isn't actually my native language.)
The issue for most creatives with AI is not “fear” of the technology, but rather resentment of a technology they believe is replacing the craftsmanship and “artistic” talent that have developed over years of labor and sacrifice.
The issue is that people are rightfully worried AI will take their jobs and their creative talents will have no economic value, in the short to medium term, they're right, but in the long term, AI will do everything and nobody will need jobs, and then artists will be able to create art purely for the enjoyment of it
I disagree, it is fear. And yes, also resentment, but that is nothing new as craftsman have been dealing with losing their livings to machines for over a hundred years.
@@PeteQuad I agree about the relationship between tech and craftsmanship over the past few centuries, but disagree about the motivation for creatives being fear. These are people that have largely been at the tip of the spear when it comes to technology and their mediums. It just isn’t a reasonable assumption that they’ve suddenly become fearful of it. Also, when asked about their thoughts, they are quite clear that they resent that people are creating tools to make their talent and discipline into binary code anyone can leverage. The added insult that generations of their art was used to train their replacement largely without consent or compensation doesn’t help.
They are not fearful. They are pissed.
@@karlwest437 That is probably how things will play out in Sweden, but in America it will not at all be that way. This technology is sure to be used to widen the gap between those with and those without. Then they will use the reaction of those without to justify some atrocious law that will codify that state of affairs and make it permanent. Our political climate when combined with our economic model make this an almost forgone conclusion.
The thinkers and visionaries have big plans, but the barbarians that used to be at the gate are now in control. If we want the world you describe, and I doubt as many as you think do, we will have to fight tooth and nail for it.
@@DeterminedPerserverance I'm not really referring to creatives in particular. I think the new generation is in general fearful of technology partly because they don't understand it. They have not seen much change in the last 15 years and the technology they use is mature and doesn't require specific knowledge.
The thing that stands out to me in that Figure demo is the tiny push of the objects right at the end, it didn't need to do that, but it chose to anyway.
14:57 This is why I value your videos the most, and it just feels more interesting and less monotone
Looks like the Terminator Movie telling us the Cyberdyne Systems Model 101 or the T-800 was created in 2024 was spot on
Matt Wolfe, you've done it again! This AI news update is absolutely essential viewing for anyone who wants to stay ahead of the curve. The way you break down complex topics into bite-sized pieces is incredible, and I feel like I'm finally starting to understand the true potential of AI. It's clear that you're passionate about sharing this knowledge with the world, and I wouldn't be surprised if you eventually launched a paid course to help people dive even deeper. 😉 Either way, I'm incredibly grateful for all the free insights you provide, and I can't wait to see what you cover next!
100% Agreed! 👌🏾
seconded,as a non techie the content is patable
Yours Truly;
- Matt Wolfe
@@nnnn-rh9eu Not true. It's just one of his bots.
Love your content. Can't imagine the information overload you deal with to keep track of all the new AI advancements...
Matt... the Friday recaps are great. Thanks!!!
No way at all this AI ends well for the majority of the people worldwide. Some will become ungodly rich and the rest will end up receiving UBI. It’s so close to their plans now and don’t think AGI isn’t already helping make decisions.
Everyone living comfortable middle class lives without having to work sounds amazing for the majority to me. Even if some rich people have space yachts.
@@MrNote-lz7lhYou’re so very optimistic… IMO everyone will NOT be living comfortable, middle class lives but instead be given just enough to survive which may include disgusting, low quality, artificial food products (e.g. made from insects). Just think about how corrupt and greedy those in power are right now. Do you believe that will suddenly change for some reason? The fundamental aspects of human behavior remain unchanged over time despite advances in tech
@DragonBiscuit
The people in power are actually pretty great people. I have full confidence that they'd give us comfortable lives of luxury. And you just have to look at historic trends to see it. Extreme poverty is on its way to extinction, and poverty has been on a sharp decline, too. All thanks to capitalism. And thanks to capitalism, robots will do our work as we live off their labor. Although I am sure a neuralink exocortex will allow you get even more by providing value to society.
@@MrNote-lz7lh Bwa ha ha ha! Sarcasm?
@@DragonBiscuit HAIL KARL SCHWAB!!!!! HAIL DAVOS!!!
I love that the robot has a good command of English. "I think I did pretty well". Most AI youtubers would say "I think I did pretty good"
An Oxford professor of English I was watching said GPT4 had a better command of prose and grammar than the majority of her university students. Not really all that surprising, even for humans the more you read and the more widely you read the higher the quality of your language. These LLMs are figuratively made of words.
I think this is the reason I found the robot creepy. it sounded considerably more intelligent than it actually was, which gave me the impression that it was playing dumb.
Devin looks really interesting. I can't wait to try it. Thanks for the update. Keep up the great work!
Knowing what details to leave out when asked “what do you see?” Is impressive. It’s intuitive for people, but hard to define.
thats why LLMS's will always rmemain dumb. Because we use language to communicte our learnings, but we dont really learn from words, welearn from sight and sound and feelings, kinda trillion times more complicated than the amount of words in a dictionary which is all it has or can possibly have.
and here is a bot with eyes and ears@@PazLeBon
@@itoibo4208 too complicated for you dimwit?
My experience is that people have a very strong instinctive negative reaction to AI. That in itself is something to consider.
Thanks for the news on the Nvidia conference. Some interesting panels although there is a lot of pay to learn content there as well. Keep up the good work.
really like your down to earth and real and common sense aproach on this topic. I just wish that youtube would also support more real and true to the fact titles for visibility
I agree, obviously, that Devin isn't an AGI. But your reasoning is faulty. AGI will likely have non-AGI components. I don't think you can take an eventual AGI system and say, well, the language reasoning module, or the math module, isn't a general intelligence, so the whole system isn't either.
Very true. It seems that LLMs will likely just be a component of an AGI instead of the whole thing.
AGI is impossible under current technology if ever. A.I is a tool that will always adhere to the wishes of it's master. As long as we don't understand what it means to be alive and human A.I might never surpass that. Humanity would need A.I to improve that and A.I would need Humanity for it's basic understanding to help us understand. We'll see if we get there. We might get there to some capacity but, is it practical? To make a totally autonomous A.I ignores the whole purpose for technology. Practicality.
@@coreym162 Well, some of these AIs show something similar to some kind of awareness by doing more than they were asked to. For example, one of them found a hidden sentence in a long test, but then suddenly all by itself told that "this sentence seemed to be put there to test me". There's a great book I Am a Strange Loop by Douglas Hofstadter where the author has some interesting ideas about the origins of consciousness and how it might arise from self-reflection. So, if someone could just run an AI endlessly and tell it "Just live and do whatever you want", we might get somewhere. However, I doubt that any company would want to burn their server resources to run an AI that just "thinks whatever it wants" without any useful input requests.
True, but in this case the LLM is the whole reasoning module, not just an isolated part of it like a math rules module could be.
The time you spent explaining that you sped the video up a little bit was very likely less then the time it took in the actual footage lol
lol... I was thinking the exact same thing. I'm chiming in for moral support. :)
I hope you had a great time here in Austin!
Matt, I'm extremely impressed with your pragmatic understanding of the Figure 1 robot. Even with the dramatic production aspect of the video, you didn't let emotions overcome your ability to process what's really going on, which helps us to wrap our own brains around it in meaningful ways. Thank you.
I personally find all this absolutely fascinating and remember humans are the creators of this magic! Like any tool it can be used for good or bad.
Hi Matt! I love your channel but was i wondering why your green screen is SOOOOO glitchy!?
awesome video, always great catching up on what's going on weekly in ai!
Can’t wait until GenAI has a knowledge window further out than 2 months in the future.
As always, and even though it is not their fault, it’s the common people fighting to keep a job to earn a living will suffer the most from these kinds of changes, until the economy and society find a new equilibrium. An economy that handles changes like these so badly needs to be replaced by something better, so that people don’t needlessly suffer its flaws.
Registered! Thanks Matt! You're the best!
Matt Wolf your going to the GTC? I am looking to go next week as well. Which events and workshops are you going to be at? I am hoping to see you there.
Claude 3 Haiku is impressively good at retrieving info from documents. I asked it to rewrite a table with over a thousand cells from a PDF document into a markdown table (not just rewriting, but combining multiple tables into one). I then spent some time checking it to see if I could find a mistake - I couldn't.
So even though it isn't even close to the intelligence of Opus, it's still very good for some use-cases.
Is Claude 3 Haiku the same as Claude3?
@@michellemonet4358 There are 3 versions of Claude 3. Haiku is by the cheapest one, Sonnet is the middle, and Opus is the state of the art. Haiku and Sonnet can be accessed for free, but Opus requires a subscription :)
I really dig the Saul Goodman Voice being used by the robot 😂
Loving AI news, great work thank you Matt.
One of the reasons I follow you is precisely because you don’t use Clickbait headlines or claim everything is AGI. It is very much appreciated. Keep up the awesome work.
Those attendees are being try-hards. It's like booing the steam engine, or the telephone. It makes no sense. But hey, let's boo something because someone else is booing.
@@VioFaxsame as your fate and my fate probably: unemployed. AI is so much bigger than past inventions. Society has been able to handle a few jobs getting automated here and there over time, but the next decade will cause a lot of disruption that the economy and society will have a hard time keeping up with. Unfortunately governments may not step in to address the issue (with an AI/robot tax and UBI) until there is a lot of social unrest. Booing is the start of that unrest.
@@ash.mysticWhich is really sad, because that's a human problem, not an AI or tech problem. If *we* can't grow up and be kind to each other, then problems will occur. But if we can, AI will be a massive boon to society and the future of the species. It'd just be nice if people put the blame in the right spot and booed our broken economic systems instead of booing AI advances.
Believe it or not, people like really enjoy seeing an AI capable of doing their career...which is more and more people each day.
Me watching Figure's Robot function: *Starts cleaning my weapons".
No need, Figure 01 can clean them for you.
Hey Matt just a heads up, your nVidia affiliate link does not work to register for GDC.
@Matt, this is not related to this video per se (you do a great job at all of them!). I wanted to share that I once saw a migraine medication being advertised to you in one of the RUclips videos you were sharing (wrap your head around that 😂). I've been suffering from migraines all my life and throughout found a few medications that work somewhat well. I decided to ask my doctor for a prescription of the med that was advertised to you (called Ubrelvy) and try it out. So far it's been absolutely fantastic and most definitely improved my quality of life. So thank you for doing what you do and positively impacting people's lives in many different (sometimes random 😅) ways! Great job!
I cannot tell you how angry that makes me that those people would boo at those peoples hard work. They came up with a creation they came up. They were blessed with an invention that can cure disease help people learn, fixing inequality fix climate change, and they do that just because those people are rich and selfish and they’re mad because they’re reign of oppression is going to come to an end with Ai replacing them and everybody’s gonna have a chance of that everyone’s gonna have a chance to get ahead and they hate it and I’m glad thank you AI.
People don't care if it's going to make them homeless.
I see AI causing a lot of joblessness in the near future. The hope of freedom is less likely in the short term.
Great the robots rounding us up as slaves are going to have the most calming psychologist voice
How will they be rounding us up as slaves?
@@michellemonet4358 it's sarcasm
For midjourney cref. Have you tried a double loop to get the best look?
Id just swap the face after generation either way to get the right fidelity in facial features.
It’s possible that reducing the weighting from the default of -cw 100 may make those specific character ref Midjourney prompts more successful. At 100 it’s trying to maintain too much of the original’s framing and content.
When AI solves the "robotics problem" they won't need humans much more.
Correct.
And now you understand why they push for nationwide installation of fast charging infrastructure.
It isn't for the EVs ... it's for their 24/7 ai-controlled enforcer robots when SHTF.
Or humans wont need the elite any more. Humans outnumber the elite. The plan might flip flop on them.
"eh edible." Does Figure One stutter?
It mimics human speech, yes.
Yes via whisper
Can you please link the Invideo link? I don't see it in your description. To watch for free.
Figure-1 seems really different, as far as I know a normal robot would fetch the apple in a similar situation starting by using the left hand and delivering it also by the same hand after analyzing the whole situation, but this looks like it's trained intelligently different. Very interesting Matt. 😎
0:30 WOW! Someone call the Ghostbusters! SXSW was full of ghosts!
He's got big hands 2:55
And you know what they say about big hands!
They need big gloves?
Well … big gloves. Obviously. Though that “other” use case is sure to be an early mover. I mean …. Weve all wanted to unplug the spouse from time to time. Imagine you can actually do that if it is a robot.
Big feet 😂
big nose
Hey Matt, looking forward to seeing some cool Shorts from GTC 😎😉
Registered for the Virtual Event through your channel couple of weeks ago 🫡
Have fun there!
Best Volker
Thank you for your honesty and not using annoying dishonest clickbait nonsense. Some of those other channels are becoming incredibly annoying.
Didnt know about the free virtual visit, this will be awesome. Thank you
One thing to note about the booing is that the dislike towards AI doesn't come only from concerns about monopolies and big tech. If you go to a few subreddits (e.g., artisthate), you see an almost murderous hate about anything AI related due to the risks it poses to certain jobs. Those people talk about and might actually start boycotting AI events since it's so easy to do. We might see this happening in every future AI event.
People who can’t handle progress will be left behind
@@xbon1 Deep down in the dirt...
Whether Midjourney is creating a good likeness of yourself based on a provided photo is probably something you should ask another person to judge. People tend to have a skewed impression of what they look like.
Ok, so I am going to give my perspective as someone who has been in the industry, and understands what SXSW is all about. A lot of people understand what SXSW is from the outside, but maybe not everyone gets it from a deeper film making perspective. SXSW is where young and up and coming talented film makers are found, its where maybe a small production studio is looking to get funded by a big company to take their movies and possibly even their careers to the next level, its where teams form that can create a movie or a short film, its where your creative idea can be shipped around to bigger players in the industry so that by the end of it, maybe you get a mention by the whole festival that has somewhat of an affect on your career. The people who are booing AI, aren't scared of the people who are controling the most foward momentum of AI, it is the fact that AI is taking away the ability to just get your foot in the door of a larger company. Its taking away the junior level/entry level jobs. Its forcing creatives to either become an AI artist and give up on the craft that they have spent most of their lives studying OR forcing them to have an even harder time trying to find funding as ALOT of big name studios are using AI to replace as many human beings as possible. You are right matt when you say that there isn't a way for this genie to go back into the box, but that doesn't really take away from the fact that almost all of the entertainment industry is getting wrecked because of AI. I mean I never would have expected AI to show up at SXSW of all places, but people shouldn't be caught off guard by people not liking the fact that AI is quite literally stealing their jobs and quite literally stealing their work and using it as their own. Its not going to be easy for anyone going forward, and Im happy to see that you are excited about the future, but for the people who ARENT excited about losing their jobs to AI, maybe give them a bit more credit because truthfully, AI wouldn't exist without the work of real artists. Real camera operators, real writers, real VFX artists, reral cinematographers, real human beings who have poured their heart and soul into this and just want a bit more credit for making AI even possible in the first place.
Thats my opinion.
I can see how another application for the videogame AI you talked about could be used to help debug games - the AI could beta test games faster and at a much larger scale, so that the computer could tell the developers areas of unexpected behaviors (bugs, glitches, etc) since now it knows how the game is supposed to go like.
Another subtle and non exaggerative video title 😂
that tube buddy is awesome!
Another great vid, there is so much new AI stuff these days it hard to keep up with it all.
That robot is completely fake. Listen to the voice. It’s clearly a man speaking saying “umm”
Plus the voice isn’t coming from the same room as the real man
You can prompt chat gpt to use natural language style. I've done this several times for immersive bots. It's not that crazy
Looks like cgi, robot can't just give plastic/toxic apple, it pushes tray too emotionally, sounds despondent as to imply it achieved emotional consciousness
Booing AI is like going back in time and booing automobile inventors for trying to kill off the horse and buggy. In the wrong hands, bad sure. But its called progress folks.
Well one could argue it's already in the wrong hands so 🤷🏿♂️
Let's boo a technology that will soon solve all our big world problems because we would have to adapt. s/
@@themplanetz everything the world comes up with gets in the wrong hands the minute anyone can get its hands on it.
Yes, because automobiles removed the last realms of validating, meaningful work for humans…right?
@@phloog Nope. It just made validating, meaningful work a whole lot easier. AI will do the same.
1:12 “you need to stop resisting” HUH???
Please do a video on the board, thats very interesting, would definitely watch it!
In the words of the famous Borg..."resistence is futile, prepare to be assimilated." 🤣🤣🤣🤣
When I try to do the midjourney thing I just get:
The content type of the provided link is text/html, instead of an accepted image type.
Why isn't it working for me?
in vision systems, color is a powerful attribute. it would be interesting to see if it distinguishes an apple without that attribute.
also it appears to be using all vision and no haptics - which will limit how it grasps things
Registered to GTC, keep up the good work once again🤩👍
I for one would actually like to know about the people from openai, seems really important to know right now and from a trusted source
Do you have a video that compares all the paid versions of: Perplexity pro to Claude 3 Opus, GPT 4 Pro?
This is absolutely mind blowing.... When you said to "watch the video", I was immediately floored and amazed. Fantastic video and I'll be watching more after this one. 😊🌎❤️🕺🏻🐶🖥️
Midjourney Pro Tip: Upload multiple photos of your character reference (front, side, there quarters view, etc...) before generating your character reference.
GPT 4.5 Turbo is already on the OpenAI API system, as I logged into the OpenAI Api platform earlier today and noticed it.
Matt - Ian Caroll made a very good video about the open AI board definitely start there
Excellent presentation thanks very much. Exciting times. James J Walsh in Limerick city Ireland 🇮🇪
Can't wait to try out Claude3 once they remove the phone number restriction.
regarding devin, it was only considering pull requests from github or something like that. not inventing the wheel but just choosing wich solution was the best. cool, but not even close to AGI.
Yes, cover all of the board members bio please 🙏🏾
Matt. I needed to hear about the booing and haters. It's important to know for sure
For the test with the robot and the dishes, unless the robot is smart enough to analyze the dishes for how clean they are, I would think it should assume that the cup and plate sitting on the table are dirty and need to be washed before putting them in the drying rack. If it were smart enough to know that they are not dirty, then it should also know that dry plates and cups don't need to go in the drying rack at all and suggest that they be put away and the same with the dry plates and cup in the drying rack.
Wondering how much action the guitar and banjo get.
22:44 i wonder how well it would do if they fed it blender tutorials. like it seams the same thing to me, fallowing what to do and copying the screen.
Nice update!
Are the robots movements controlled by gpt4 or is this different software?
Who remembers the previous video in which a person appears teaching the artificial intelligence how to move and do household chores, and the person moved it every time to learn? I think this is the final result after all those movements, but it is limited to the movements that it has learned.
Great News as always
Thanks for calling out the overuse of "AGI", I find it really obnoxious and, while your thumbnails/titles are a LITTLE clickbaity, I can usually rely on you to give sort of hype/buzzword free, logical assessments of new technology, which is amazingly valuable. Thanks for doing all that you do.
One has to be clickbaity, state of the biz, clickbacit is literally one of the few options people have to amplify messages
Curious Ai bro sir MW would you def pay for the new Claude Pro over Chat Pro now?
Wow the robot had a nice smooth movement n touch w the fingers like we’ve never seen indeed
Waiting for world first AI phone . Fully automated , talking with you naturally, helps you to sleep better . Prepare your days , protocols and so on
Honestly, the world is pretty much the same as it was since the last half-dozen world-changing AI events came and went. Right now, AI reminds me of the Segway...and we all know how earth-changing that was!