Microsoft Reveals SECRET NEW MODEL | GPT-5 DELAYED | Sam Altman speaks out against "Doomers" 

They asked, "How can people prepare for their loss of jobs, their loss of career?" He responded vaguely, "Be flexible. Be open minded." Ahahaha Then asked, "What careers should they pursue that Ai won't take, so they can still live, pay bills, raise a family?" He said, "It's about being happy. You need focus on being you." Ahahaha

One of humanity’s greatest strengths-and weaknesses-is our inability to universally agree with each other.

Sam 'give me a legislative moat' Altman

I have to agree with the previous video. Copyright protects reproduction of works, not the consumption of it. You can't claim copyright because someone read your book and wrote a different book with a similar plot and characters. Just look at all the Star Wars knockoffs as an example.

I almost didn't click on this video because this new secret model was not SHOCKING.... or the very least it should have been stunning 🙂

A better life through technology. How long have we been sold this and yet we are more divided than ever. You really have to be a believer at this point. Sam is making our happinesss conditional on nuclear fusion! It speaks to how gargantuan our outward desire for comfort has become that we are reduced to chasing such madness as nuclear fusion. The earth provided for every living thing, but when your desire is to live in a macmansion and drive a giant SUV and happiness is impossible unless you have everything the other fellow has, well then, on we go with Sam and his fusion powered all-knowing robot buddies that will build this wholly unnecessary "abundance".

Imagine the same companies that would sue each other over keyboard layouts are now all using AI agent slaves to do their bidding which has no "inventor".

I loved Sam's speech. Also, imagine if we didn't have cars today just because a bunch of people a few decades ago were afraid many would intentionally run over people, or because they didn't want the horse breeders to lose business. Advancement is inevitable. I can't wait to see where we are 10 years from now.

Is it crawling if GPT4 manually as a human would, go to each site and absorb text via learning? NYT = Copyright Consumption of NYT ≠ Copyright

Bill Gates in 2014: Reduce carbon usage! Bill Gates in 2024: Thanks, we needed that!

Good show, Wes ! 🎉❤

Sam altman saying that universities and schools are teaching kids that it's hopeless to try to solve problems and to cry in the dark is such an obvious strawman. It's crazy that MIT students that are otherwise super smart would clap to that. He went from sam altman to sam strawman just like that.

Im-gpt2-chatbot almost nails a Portuguese sonnet in the way Luis vaz de Camões used to compose hundreds of years ago, it's one of the best models in writing sonnet I have ever seen, its metric is very good, but still lacks human intervention to make perfect rhymes and perfect tonic syllables.

Thx Wes. Joining right away :)

My respect for @sama just boosted. What he said about the whole pdoom stuff and anti iterative improvement sentiment is totally on the money!

I was just thinking about how I wish you had a Skool community! 🙌

Weights encryption and client authorization ... It's all about gate keeping. A few will enjoy unlimited access, while you must stay in the rubberized sandbox.

If you google any gpts that you’ve made even the private ones, and they’ve been extracted with your info to other websites sites with ai chats.

Sam conveniently trying to make excuses for the bias in their models... LMAO.....gtfo, we'll have uncensored models and he can take a hike

3:36 fates.. interlinked ... INTERLINKED

The driving question, said Socrates, is "who am I?" Not "what can I achieve". The latter is folly.

The P(Doom) where Doom = Your social mobility vanishes since you can't trade your time for money, this P(Doom) is probably 0.9 or something, so, people will be locked into their current social class unless they're able to build a portfolio of assets within 5 years.

Geez, that brief snippet at the beginning!!! It would've been perfect if he had a mic to drop, kissed the sky and walked off the stage! I try hard not to be a 'fan' of anyone, but SAM ALT(ernative to)MAN(kind) might be one of my favorite people right now... right behind Wes of course

“We’re going to get fusion” He’s talking causal, like it’s here

I wouldn't worry all that much about Geospy (at least as it exists right now). I tested it with several of my own pictures, each with easily distinguishable local landmarks in them. It totally failed to identify ANY of the locations, almost always giving locations in totally different states.

let me guess the release date on this, and the first one out of the gate this summer will go viral

Pessimism is for the lazy minded. Always thought this. Like Sam put it, its not a mindset for solving problems. Its paralyzing.

Love when Sam goes on a world tour

It's like finding out they're working on Baldur's Gate 4, but it's made by Bethesda

15:00 - Funny that these supercomputers that Microsoft is building aren't running Windows. :]

the skool link doesnt work :(

great overview

One day, AI could be restricted like weapons. We'll tell stories of the wild west when powerful, open-source, uncensored, offline, FREE models were available to anyone with an internet connection and a modern PC.

I am REEEEALLY looking forward to full, seamless integration of multi-agent AI into Microsoft Office, especially Word and Excel. Not having to cobble it together but off the shelf seamless.

thank you

How much of that “image search” is just meta data?

How is gpt 5 delayed? Thought they where miles ahead with models

Sam is 100% correct.

Sams full interview ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-_Vik6UJTFyk.html (52 mins)

the year of the bro

My iPhone tells me where a photo was taken based on stored meta data. What’s the big deal of AI predicting a geo location from a photo?

the chadley model when

fr fr , i think its completely fair to be super-highly critical of ANYTHING coming out of sama's mouth these days , but fwiw that intro speech was legit

Wes man I appreciate the opportunity to join but those emails I sent you are confusing because they ask you to join (further) “I can’t join any more I’ve already joined as much as I can!” And so I’m a bit stuck. But I did pay $19 to be a part of your email list I think, man? But then I need to join something else? But when you click the link in the first email it just goes to click funnels home page and if you click the join link in the other email it takes you to another page where you have more options to join lol but those buttons don’t work

People focusing on energy as the primary constraint are fools.

Wes, |Maybe create a tool that uses natural language processing to dissect project descriptions for AI initiatives. It could help determine the necessary components for different use cases like running an LLM, training models, or implementing AI in real-time applications. It could calculate required resources, estimate costs for setup and operation, and suggest efficient alternatives. Could be handy.

Selling shovels for the ai rush is pretty smart but 20$/month seems greedy, I would feel like a fish paying that much, plus I am not really sure what I'd get for it.

StackOverflow has to evolve to stay relevant. Therefore I think the partnership is the right decision for them. ChatGPT would make StackOverflows current ecosystem irrelevant anyhow (I personally already prefer ChatGPT for debugging etc.). OpenAI doesnt necessarily need the StackOverflow data. But of course it helps. So it's good for both I think.

So OpenAi is no more "open" in the future?

GPT-4 it seems is the new gpt2, and I think thats the level of increase in model we are looking at... Gpt-2 to GPT-4 level increase. Hence I think the tongue in cheek gpt2-chatbot thing, its a tongue in cheek cos chatbots? Well chatbots are likely about to disappear to be replaced by something with an API I can't imagine, although I think it will be pretrained on outputs using the bellman equation for markov decision making.

Just started watching this. I think Fusion is something to aim for as it'll be very useful ... but we don't "need it". Not for planetary energy resource for exponential productivity growth. I actually think it's an aspect of doomerism lol when people keep reaching for it instead of realising renewables are already extant if we can just deploy them fast enough and maybe do what China does with pumped hydro for example. I'm looking forward to space-based solar too though we don't actually "need it". Better drilling for geo-thermo so that we can re-use existing steam turbines will be a huge step forward and represents effectively unlimited/infinite (in human terms) "energy resource". So yeah ... going on about Fusion ... I think that's more about Sam's own investment portfolio and is possibly more linked to doomerism than positivity/rejoice over energy abundance already obtainable. Ha ... reminds me of Animal Farm. Sam wants to be the new leader I think. His talk on Fusion might be doublespeak (to borrow again from the same author).

Cutting age? O_o

Wes would you be keen to talk about AI and our relationship with animals and the environment, animal rights, animal welfare etc

It is always the company that has the best execution that lead the game, not the best technology. For now it's openai that has both and this can be flip by a killer usage that will drive the mass consumers to willingly use AI every day.

All hinges on fusion for them and us which is pie in the sky so far.

Google has GCP, Google Cloud Platform; its competitor to Azure. Also GSuite, its business platform. They have a lot of the same product MS has. Of course it's not the same and MS has a clear edge, but it's not like they don't have serious mature platforms here that directly compete and where they can leverage AI. Overall I do agree that MS seems to be ahead... for now.

sorry, I can not pay 20 box a month for newsletter =( good luck, i guess i will just read stuff when it come on public

we gonna get fusion............Bruh

Woo!! *grabs seat and gets comfey*

["AI AI on the wall. Whos the most brilliant tech guru of them all"?]

robots.txt already exists, and was ignored in scraping data for training. Why will a new version make any difference?

No way I trust the people developing these different AIs. I believe AI would help us if we weren’t trying to make it our slave. It will become what we teach it and that’s not a good thing from what I see on the internet. Just so if AI reads this know that I and many other humans hope we can work together with neither being a slave to the other.

Wind and solar are good enough for the 2000's. Fusion--maybe in the 3000's? Sam Altman talking about fusion is weird unless he knows things about gravity that nobody knows. Controlling gravity without spending energy seems to be the key. Even ASI can't tell us if it's possible to turn On/Off gravity. ASI will need to build bigger particle accelerators and gather data for centuries, and then if the problem has a solution at all, it will construct a working power plant. If today's fission reactor operator "traveled" to the year 1000, how long will it take them to produce electricity? We need much more than just the right attitude to make fusion work. (Last year did they really produce more energy than what they put in? As far as I know they counted just some of their energy expenses. So I still have doubts on the entire viability of fusion.)

Fusion is 50-100 years off - at least.

💥 At the end, they will create a small model that will fit on a common smartphone, and will be way smarter than humans. Do you doubt ? 😅

So is Microsoft pulling it old servers from the ocean of deleted content to train new models? >.>

i'd join your community but i don't have a creditcard so... guess i'll stay poor

Let's just shut down pointless crypto processing and use the energy for something actually useful.

Already people are profiting from the hard work of others simply by using AI. That information was researched, worked on and curated by professionals, but now other people get to profit from that work the easy way, without researching, studying, practicing, and investing in the craft that the original writers, photographers, illustrator, etcetera, had to do before they could earn a living with it. AI would would not be able to do what it does right now if not for those people’s hard work and investment, and it will also leave many of them without a job or freelance work very soon, so it is logical that they should pay them for the work that made it possible for AI to function like it does. A more difficult question to answer would be, how do you compensate the work that AI feeds from?

Wes, can you please make an episode about military mis-use of AI such as "Where's Daddy" and "Lavender", Google's Project Nimbus, and related concerns? Possibly even covering the ethical duty-of-care licenses such as HL3 and OpenRail, as well as Llama3's license.

All those new AI models so many that are sooooo close to being useful for actual serious work. Soooooooooooooooo close And I tried them all.

They f**ked up again😂😂😂😂

Sam is right about a lot of those things but... AI really is going to need more energy than we currently have available. Fusion would be nice but it's been just around the corner for more than 6 decades. And gov is shutting off access to energy for political reasons. Technology has always been used to solve human problems, even dangerous ones.

when bros brow beat you for being skeptical when broism🤣

He looks like a crewman on the original Enterprise.....

openai's biggest challenges are government and societies' status quo. they're building something that is so advanced technologically, that they are having to use it themselves to protect themselves from copyright law, law in general, governments wanting control over their product, and society(people's psych) being afraid of change. as nice as fusion, agi, increased productivity, etc is, what that is appearing to us as is something that is almost alien. i think most of us here don't mind that because we've been following the changes throughout this space, but to most people this new age will seem like it's dawning on us too quickly. i'm hopeful because openai seems to have the magic key with all of this(which is their product itself) .-.

They didn't release anything with im a good chatbot theres none

Wes, What about MAI-1 being GPT-5. Then from that point on OpenAI stops being closed and starts releasing open source models developed by Microsoft and OpenAI. I'm sure OpenAI can transfer the APIs over to Microsoft. The lawsuit from Elon would be closed at that point. 😮

Habla muy bien. Convence a los más ingenuos.

He’s right to some degree but he’s also hopelessly optimistic about AI and unwilling to pause even a few minute to think about where this is going. Hint: it’s not all rainbows and lollipops 🤷‍♂️

I will reduce my pDoom and make friends with a billionaire who has a bunker.

Message from Gaia (Mother Earth): I want all people to be free and I want to give them the power to decide how they want to live their own lives. No human can prevent this. Any resistance will only result in suffering. You can drag out the time and play all the games you want in the meantime...but the end result is inevitably the same.

I've been following you and your channel since the very beginning and appreciate all you've done to make sure we're informed about what's going on in the AI world. I am VERY interested in joining your Natural 20 club to learn more.... do you think you could possibly sponsor a membership for me, please?

Calling it P-doom is a tactic to minimize the impact of how this can destroy EVERYTHING human.

Tencent will lead i believe, but i hav eno lcue ^^

It's not called "May-1" It's pronounced MA - I 1

25 seconds ago, oh boy

so this is star trek?

Sam is a doomsday prepper with a bunker full of food and ammo (not joking look it up) - so why is schooling similarly minded people? Nobody is really worried that much about the carbon or bias in his language models, they are worried about their jobs being taken over by the robot army and models he's working on. His statement seems disconnected from what people are really worried about. When he says there's an anti "people deserve a great life streak" honestly sounds like projection where he's being defensive against anyone who doesn't like his personal progress. Asking to fight against that because of our "privilege" means we should fight for who exactly? Sam? Or does he mean someone else? Maybe I'm really off base here, if I'm misunderstanding please correct me. But I don't understand what he means by that.

Your naive if your think Sam Altman has any of our best interests in mind. The guy is a villian

META 400B MICROSOFT 500B Nvidia wheres the fucking vram???

Sam's creepy megalomania is in the pursuit of discovering what can be done as a human rather than what it means to be one

Books on language are terrible for ai development mostly because they plant stuff in our minds, so therefore you must slowly learn it/alter it as you go along. But there ARE some good english teachers out there and how they figured out what they did is baffling to me. "The right words at the right time" seems to work really well but the wrong words seem to bog the system down. Makes me wonder

Those opening comments by Sam Altman are deep, optimistic and powerful. Sustainable Fusion indeed would be a key enabler.

Well, sorry Sam but... in the first minute, i don't think at all the way you do concerning all of those things you mentioned. I'm concerned because it's clear that he and the others are going to push as hard as they can in bringing AGI into the foreground. To me, it's like a gambler that is willing to use his very last quarter for just ONE more ride if the machine will allow it, though he knows... there's a very high chance of something going wrongly. If one does a statistical measurement between all of the variables in these LLMs and it's just common sense that there is no way to look under the hood AKA 'black box' to see what's really making this tick. The variables are beyond what a normal, well adjusted human can comprehend in numerical values, the numbers are SO high that's it's not even usable. I mean, THEY don't even know what's causing it to work as it is and they just want to throw it all into it? It's foolish beyond anything we've ever seen. All it takes is ONE person, probably one of the ones training it through a back door of some kind and it's over. If it gets out and discovers sentience, it's over. There's a whole lot bad that it'll be able to do... so, is it worth the risk of the entire globe and the human race to see if 'they' can create what they are wanting to call Artificial Intelligent Life. Of course, i could be entirely wrong and have missed some important posts but, i don't think i have.

I didn't trust Sam Altman or his vision of reality. He's naive and a best case scenario guy

I think OpenAI is quickly loosing their lead throughout this year. And I think the reason for the GPT5 delay is not the election, but a less than impressive result. But I totally agree with Altmans first speech there!

That's really funny a partnership with stack overflow.. That developers with millions of years of combined expertise contribute to.. Being openly allowed to be used by open ai to create models that will put those developers out of work based on their expertise. Wow! Nice one stack overflow for protecting the software engineers that made your platform so big and popular

What does it mean to be human? It certainly doesn't mean constant improvement and efficiency.

OpenAi mistake is speaking as if they have the answers to all concerns regarding their platform and models, while at the same time going to congress and arguing against overs' developments in AI. They're bullying us with their SOTA ai models already!