I wish I was younger than 73 to watch this AI rocket ship take off. I was in the bay area selling 8080 systems and using Windows 2.0 for thousands of dollars back in the 80s. My wife was international controller and an engineer approached here with a chip design that would allow everyone to have a phone in their pocket. After doing the ROI they both made a presentation in the new products meeting explaining the opportunity. It was decided to pass because there were phone booths on every corner. The engineer went to Motorola and the rest is history. Hang on folks the next pardine shift is here !!!
This feels like a trojan horse for absolute control over humanity. We don't need AI controlled by a select few, we need personal AI that is secure and free from outside control.
100%. It's scary how people like Larry E. From Oracle is shaping how AI to be used... Please go and see his comments about full surveillance of society using AI... we need to push against that before it's too late.
Impossible. The amount of power required is too large to not be collaborative. The best we cn hope for is an unjudgemental uncensored AI that can consolodate all data. Maybe we could have a true democracy in the future with a brain/computer interface.
Getting 175th in a programming competition has almost nothing to do with being the best programmer the best programmer is a great team member, can anticipate and even suggest a future vision for a product, is active in requirements gathering and will sometimes push back if the given task is not in the best interest of the organization, and if we incorporate in the meaning the best programmer for society at large they are also considerate of who they work for and what they do for organizations, programming competitions have virtually nothing to do with any of those traits.
Yes, that’s why openAI has Stage4 Innovators (like you mentioned) and Stage5 Organizations , on their roadmap after Stage3 Agents. We are at Stage2 Reasoners sort of. Somewhat underwhelming since not true reasoning like you mentioned. But we will get there in a couple years.
this is absurd,it takes time to build infrastructure.power availability,fiber access,real estate...these alone can take years ...Oracle doesn't have no scale expertise,they are miles behind amazon,microsoft,and even google.
Before 2027… ppl need to start preparing. No way UBI will be ready by then. Shaping up to be some really rough years before things hopefully gets better😮
Just prompt it to work for you and be an agent. “Super agents” coming. If these things are actually truly smart and autonomous, it’ll be able to make EVERYONE money! Otherwise it’s corrupt and doesn’t make sense. UBI isn’t necessary if technology becomes truly useful.
... a joint venture company to build data centers to train and run the inference for large language models in development currently and in the future by the respective partner corporations that constitute the joint venture. It's a company that runs data centers. It is not that complicated.
Quantum computing is still unstable and theoretical. At least this technology reaches the public more. Normal people can’t use quantum computers atm. Even if we invest 500 billion…that might not return anything. But who knows.
This was a power play by OpenAI because they're aware that AI in general plateaus in functionality and usecase after a while. There is no moat, and so opensource models are catching up quickly (see deepseak-R1). Their latest model that cost $200 per month barely does better than free models. So rather than get better or more innovative, they took the Microsoft approach to just cement/supplant themselves everywhere in our echo system. I see it as a net benefit if we get AI at say the DMV and other local government agencies (or the post office) to improve quality of life.
What about the unknowns? Even now we "don't understand" why AI hallucinates. AI is not a new toy. It is a revolution even we don't completely understand. Being glib about the future with rose-colored glasses is premature.
that's why you should invest more $$$ into AI companies - to make better AI with fewer errors :) Also, there's Ai companies (like Recursion Pharmaceuticals) working on making drugs to treat diseases or perhaps cure diseases one day. How cool would it be to have Ai doctors that ask you for your symptoms and use Reason and their encyclopedic medical knowledge to diagnose health problems? Also, Ai girlfriends - not my thing but it could prevent lonely young adults from shooting up schools. Downside of Ai: - could be used by military to create advanced weaponry - decimate customer service industry jobs and jobs in other sectors
No discussion about the environmental impact of these exponentially growing models. The 'bigger is better' paradigm doesn't seem to me to be sustainable. Pumping half a billion dollars into the process may yield a result, but at what cost?
Thank you, OpenAI, for giving me a tool that enhances my personal autonomy, empowers my decision-making, and ultimately contributes to my self-sovereignty! (Yes, I did use co-pilot to help write this comment)
Here are some ways AI can help: 1. **Personalized Learning**: AI can recommend educational resources, courses, and content tailored to your interests and needs, helping you expand your knowledge and skills. 2. **Financial Management**: AI-powered financial tools can help you budget, invest, and plan for the future, giving you greater control over your financial well-being. 3. **Health and Wellness**: AI can provide personalized health advice, track your fitness goals, and even offer mental health support through apps and virtual assistants. 4. **Time Management**: AI can help you organize your schedule, set reminders, and prioritize tasks, making you more productive and efficient. 5. **Creative Assistance**: AI can assist you in brainstorming ideas, writing, creating art, and other creative endeavors, enhancing your ability to express yourself. 6. **Decision-Making**: AI can analyze data and provide insights to help you make informed decisions in various aspects of your life, from career choices to personal relationships. 7. **Privacy and Security**: AI can help protect your personal information and online activities, giving you more control over your digital presence. 8. **Self-Reflection and Growth**: AI can provide tools for self-assessment and personal development, helping you set and achieve your goals.
So, the taxpayers will pay 500 billion and still have to pay for the tool? At what point is this going to help the people? If your AI tool is as good as you claim, you should be able to pull yourselves up by your bootstraps like every other sorry sucker who pays taxes.
... Why are the people in these crowds pretending like using these LLM's currently is particularly useful? It feels like there's a massive disconnect between people like this, who seem deeply impressed with these models current capabilities, and normal consumers and professional who use LLM and find them woefully inadequate for any of the tasks set out for them. Relying on these LLM's, currently, just degrades the quality of your final output, in exchange for a moderate increase in the speed of production -- and that's after becoming intimately familiar with the models and their usage and prompting skills, so as to best harness their potential. Frankly, I have yet to encounter a single moment with any of these LLM's where it's done narrative design that I would even find acceptable from a new hire developing a mod for a game as their onboarding practice -- nevermind as the actual narrative design for an implemented deliverable in a releasable product. Maybe Stargate is what these models need to finally become more competent than a 16 year old, but at the moment they're basically just replacements for personal assistants and executive assistants -- of middling quality.
Imagine Stargate hallucinating on healthcare and pushing for the wrong mRNA to have our cells code foreign proteins. Stargate needs to gatekeep away villainous power hungry execs from building a modern day Tower of Babel.
I wish I was younger than 73 to watch this AI rocket ship take off. I was in the bay area selling 8080 systems and using Windows 2.0 for thousands of dollars back in the 80s. My wife was international controller and an engineer approached here with a chip design that would allow everyone to have a phone in their pocket. After doing the ROI they both made a presentation in the new products meeting explaining the opportunity. It was decided to pass because there were phone booths on every corner. The engineer went to Motorola and the rest is history. Hang on folks the next pardine shift is here !!!
I'm 70 and I know exactly what you mean!
This feels like a trojan horse for absolute control over humanity. We don't need AI controlled by a select few, we need personal AI that is secure and free from outside control.
Then build it for yourself, otherwise, how will that work? If it's given to you, it comes from "outside".
@@dougchampion8084 he is talking about true Open-source AI what deepseek and others are doing with the MIT license
100%. It's scary how people like Larry E. From Oracle is shaping how AI to be used... Please go and see his comments about full surveillance of society using AI... we need to push against that before it's too late.
One word: Skynet
Impossible. The amount of power required is too large to not be collaborative. The best we cn hope for is an unjudgemental uncensored AI that can consolodate all data. Maybe we could have a true democracy in the future with a brain/computer interface.
The last point about being able to visualize exponentiality is the most important because basically everything is exponential at this point
Prefer these interviews over Apple interviews already haha
AI is your time saver. Period.
over the period of time Brain tasks are going to shift drastically
Getting 175th in a programming competition has almost nothing to do with being the best programmer the best programmer is a great team member, can anticipate and even suggest a future vision for a product, is active in requirements gathering and will sometimes push back if the given task is not in the best interest of the organization, and if we incorporate in the meaning the best programmer for society at large they are also considerate of who they work for and what they do for organizations, programming competitions have virtually nothing to do with any of those traits.
Yes, that’s why openAI has Stage4 Innovators (like you mentioned) and Stage5 Organizations , on their roadmap after Stage3 Agents.
We are at Stage2 Reasoners sort of. Somewhat underwhelming since not true reasoning like you mentioned. But we will get there in a couple years.
It would have been nice if he talked a bit about DeepSeek
What I do remember is when the internet was getting built out, there was overshoot in laying fiber and there is still a lot of fiber that is dark.
I used chatgpt to summarize the transcript😉
Give the chat gpt answer
This dude's sideburns go halfway up his head
I am equal parts excited about the development of ASI (o3) and anxious about the national and global state of democracy.
Can't believe she didn't ask about GPT-5
They're currently training o-4 and o-5 is supposed to be AGI.
They should drop the “Open” part of their name since they're so cagey about details.
Man these dudes don't give clear answers on anything lol
Ya gotta have money to do stuff that matters to you. AI is already displacing people from their jobs. 😢
Release full o3 and GPT 5 please! Can't wait to see o4 now lol
Contrast this to Dario... OpenAI is a company full of sleazeballs
Yeap, somebody was complaining in comments that Dario stutters and they need a good public speaker. Thank you very much.
Models get cheaper but my subscription remains the same😢
They're still losing money on this....
They get cheaper and better
What is it with soccer forms? Apparently the Singularity and possible extinction of all life on the planet are better than filling out soccer forms.
Hahaha - OpenAI is seemingly complacent with ethical and even lawful risks. I believe it will be their downfall.
this is ridiculous. some words on the screen won't kill anyone
we're all gonna die
The theft of our energy, I didn't ask for that
Your issue is with capitalism, not these specific companies. Act accordingly.
this is absurd,it takes time to build infrastructure.power availability,fiber access,real estate...these alone can take years ...Oracle doesn't have no scale expertise,they are miles behind amazon,microsoft,and even google.
Stargate is a big deal!
Wsj is trying real news... wow
Before 2027… ppl need to start preparing. No way UBI will be ready by then. Shaping up to be some really rough years before things hopefully gets better😮
Just prompt it to work for you and be an agent. “Super agents” coming. If these things are actually truly smart and autonomous, it’ll be able to make EVERYONE money! Otherwise it’s corrupt and doesn’t make sense.
UBI isn’t necessary if technology becomes truly useful.
Why did trump announce it? What is the US investment into this?
It’s going to be partnering with our healthcare infrastructure data systems
Yep
Terrifying
I’ve watched several interviews. Still yet, no one knows what Stargate is. Something smells like surstromming
It's a AI datacenter + electrical infrastructure.
... a joint venture company to build data centers to train and run the inference for large language models in development currently and in the future by the respective partner corporations that constitute the joint venture.
It's a company that runs data centers. It is not that complicated.
If I owned stock in GOOGLE, I’d be worried
Or, Chat-Jippity, as ThePrimagen calls it
Deep Seek is already there which Kevin claims O3 to be.
And jack dorsey?
200 USD/ month is higher than Mercedes's monthly leasing
No
The only purpose is to keep attracting $$$. There is no moat.
"ClosedAI Product Chief"
Don't need Bill Gates within this program
Hope it works on Linux
It’s not running on your laptop bro
@ avoiding the copy paste I thought was. Claude can install on windows
Magic beans, please invest!
Got to have power to operate these massive data centers. It would be smarter to sink 500B$ into the development of quantum computing.
Quantum computing is still unstable and theoretical. At least this technology reaches the public more. Normal people can’t use quantum computers atm. Even if we invest 500 billion…that might not return anything. But who knows.
This was a power play by OpenAI because they're aware that AI in general plateaus in functionality and usecase after a while. There is no moat, and so opensource models are catching up quickly (see deepseak-R1).
Their latest model that cost $200 per month barely does better than free models.
So rather than get better or more innovative, they took the Microsoft approach to just cement/supplant themselves everywhere in our echo system.
I see it as a net benefit if we get AI at say the DMV and other local government agencies (or the post office) to improve quality of life.
It seems Not making money for them😮
9:40 - that was the best example of women hearing everything that might have been said, that I have seen in years.
Claude is better
Anti China comments as justification to do what you want makes this Harvard guy seems like a moron.
Great interview.
He really fills your heart with life affirming feelings ♥️
Is it just me? This guy has a wacky hair.... Did AI cut his hair? Something is wrong with him after Facebook and now OpenAI.
F AI!!!!!!!
so glad Trump is backing Open AI
Babylon
Compelling speaker who needs to find a different barber.
What about the unknowns? Even now we "don't understand" why AI hallucinates. AI is not a new toy. It is a revolution even we don't completely understand. Being glib about the future with rose-colored glasses is premature.
that's why you should invest more $$$ into AI companies - to make better AI with fewer errors :)
Also, there's Ai companies (like Recursion Pharmaceuticals) working on making drugs to treat diseases or perhaps cure diseases one day.
How cool would it be to have Ai doctors that ask you for your symptoms and use Reason and their encyclopedic medical knowledge to diagnose health problems?
Also, Ai girlfriends - not my thing but it could prevent lonely young adults from shooting up schools.
Downside of Ai:
- could be used by military to create advanced weaponry
- decimate customer service industry jobs and jobs in other sectors
i can sense the next luigi mangione coming for any of them
I love how he hasn’t answered a single question im five minutes in- horrible “interview”
No discussion about the environmental impact of these exponentially growing models. The 'bigger is better' paradigm doesn't seem to me to be sustainable. Pumping half a billion dollars into the process may yield a result, but at what cost?
Up to 100 BILLION in taxpayer's money that will please the Oligarchs.
Sorry FEMA is broke, live in a cardboard box.
Welcome to Trumpistan!
Thank you, OpenAI, for giving me a tool that enhances my personal autonomy, empowers my decision-making, and ultimately contributes to my self-sovereignty! (Yes, I did use co-pilot to help write this comment)
Here are some ways AI can help:
1. **Personalized Learning**: AI can recommend educational resources, courses, and content tailored to your interests and needs, helping you expand your knowledge and skills.
2. **Financial Management**: AI-powered financial tools can help you budget, invest, and plan for the future, giving you greater control over your financial well-being.
3. **Health and Wellness**: AI can provide personalized health advice, track your fitness goals, and even offer mental health support through apps and virtual assistants.
4. **Time Management**: AI can help you organize your schedule, set reminders, and prioritize tasks, making you more productive and efficient.
5. **Creative Assistance**: AI can assist you in brainstorming ideas, writing, creating art, and other creative endeavors, enhancing your ability to express yourself.
6. **Decision-Making**: AI can analyze data and provide insights to help you make informed decisions in various aspects of your life, from career choices to personal relationships.
7. **Privacy and Security**: AI can help protect your personal information and online activities, giving you more control over your digital presence.
8. **Self-Reflection and Growth**: AI can provide tools for self-assessment and personal development, helping you set and achieve your goals.
if you close your eyes he sounds like Elon Musk
Claude Sonnet 3.5 still rules.
f-ai lol
🤖🤔🎶🙌
😁
This is why we don't have free healthcare.
So, the taxpayers will pay 500 billion and still have to pay for the tool? At what point is this going to help the people? If your AI tool is as good as you claim, you should be able to pull yourselves up by your bootstraps like every other sorry sucker who pays taxes.
It isn't taxpayer funded. It is a privately funded joint venture.
WSJ interview flirting instead of holding this weirdo accountable
... Why are the people in these crowds pretending like using these LLM's currently is particularly useful? It feels like there's a massive disconnect between people like this, who seem deeply impressed with these models current capabilities, and normal consumers and professional who use LLM and find them woefully inadequate for any of the tasks set out for them. Relying on these LLM's, currently, just degrades the quality of your final output, in exchange for a moderate increase in the speed of production -- and that's after becoming intimately familiar with the models and their usage and prompting skills, so as to best harness their potential.
Frankly, I have yet to encounter a single moment with any of these LLM's where it's done narrative design that I would even find acceptable from a new hire developing a mod for a game as their onboarding practice -- nevermind as the actual narrative design for an implemented deliverable in a releasable product.
Maybe Stargate is what these models need to finally become more competent than a 16 year old, but at the moment they're basically just replacements for personal assistants and executive assistants -- of middling quality.
it's you, you're the problem
I bet you all didn't know that you can start balding from above the ear huh?
He obviously finds her attractive
First 🎉
Imagine Stargate hallucinating on healthcare and pushing for the wrong mRNA to have our cells code foreign proteins. Stargate needs to gatekeep away villainous power hungry execs from building a modern day Tower of Babel.
there has been ongoing research from before you'd heard about GPTs, multiple recent breakthroughs, just on hallucinations
I think that's still a far bit away. The immediate use will be in increasing claim denials...
So far every human born has died
Tower of Babel was precisely my thought.
This is a defence project of course - they could have called it Skynet
The country is how many trillions in debt and the government is committing another half trillion?
Government isn’t spending anything it’s the companies that are.
I want to give the AI control my computer and all the apps on my computer and I want to train it how to do eBay listings for me