Choosing a concept space to stake your next software development project on or create a startup around these days feels like trying to build a sand castle on a beach being hit by a tsunami.
Yes and no. I think the principles he described ("don't try and plug gaps in models, but build specialized products that get better as models get better" are relatively straightforward to follow.
This is true. Yes OpenAI is a model company. You have to be far away from AI to win. But also have to have critical speed because OAI is working outward to every other space. Say you are a food delivery service and you want to use AI. It can help you until there is a consolidation of food delivery that uses the best model that improves cost, speed and quality. To the sand analogy. Build it near the boardwalk and hope the water doesn’t get close.
He says, "Don't build a patch for the current models" and you'll be okay. I'd like to know if ChatGPT Search and Custom GPTs are "patches". The truth is that anything that their models can do, they reserve the right to do for themselves.
The more I listen, the more this sounds like nonsense. "Bet on the models getting better." We are talking about ARTIFICIAL INTELLIGENCE. Intelligence literally spans everything. So there is no limit to what "better" includes. A model that can drive a car, run a country, colonize Mars.... There is nothing out of scope for a model that simulates human intelligence.
Glad to hear the possibility of no-code tools empowering non-technical founders. I am building Licode, a no-code platform for building AI apps. Excited to see where it can go. :)
It could easily make that startup obsolete as well. But i guess we could always use the model itself to innovate and create new things. Im building a language with an interesting idea with the help of AI. Im moving with super human speed and i don't have any coding experience.
In the bin unless you are from a rich family with a history of committing sexual assault. Face facts. There hasn’t been anyone in this space who didn’t already start out rich for 20 years. Time to forget the startup lie and build and sell a product that works to someone that needs it.
Sad but true: “There's a lot of people in the world that I'm sure would be phenomenal AI researchers had their life paths just gone a little bit differently”
🇧🇷🇧🇷🇧🇷🇧🇷👏🏻, Sam Altman is an exceptionally gifted leader with a sharp, visionary perspective that few possess. His keen sense of direction and strategic insight set him apart as a truly unique visionary. He's gifted that way!
Interesting interview. A bit over obsessed with "being in trouble for this one" - the point of the interview is to ask probing interesting questions, it was funny the first time but got old!
@@RusuTraianCristian yeah like naming company Open.. saying he doesn't have equity there (while he has options which is same thing).. this guy is sociopath
Great interview, Harry! Sam seems very smart, though his responses feel a bit like incoherent based assumptions rather than clear strategic reasoning, but maybe that’s just my impression.
6:06 Rather Build something to go Deliver the great AI tutor or the great AI medical advisor. 32:44 AI Lawyer, AI CAD Engineer 16:44 one of the Retirement thing I fantasies about doing is writing a book of everything I have learnt of how to build an organization & a culture that does this thing ( Research Lab ) not an organization that copies what everybody else has done. 23:39 Anthropic model for coding? 34:35 Cursor
I like the last question, I tend to think the same as Sam, it's one thing to improve the tech but it's another to change the world. Especially the government, I think it will take time for these technologies to be widespread. I mean it's one thing to have CoPilot helping you to write an email. Or using UiPath RPA to automate some tasks. But it's another to replace entire jobs/teams with AIs.
Completely aligned with Human tendency to overestimate Speed and underestimate Scope of Societal Change enforced by Disruptive Technologies. Valid in the start of the Technosphere and Digisphere, valid in the start of the Cognosphere.
Excellent interview. Sam shared some helpful insights. there’s tons of opportunity to apply powerful AI agents to benefit the world in many verticals. For example, ensuring people aren’t exposed to toxins from plastics (such as BPAs from food storage, etc.) and many similar products that we all use everyday. There are many of these kinds of issues, and an AI can probably help identify the most important ones to work on first.
26:39 the heaviest thing in life is node_modules, Harry. It's node_modules. We should move on from JavaScript instead of pushing AGI agenda , agentic worlds, yada yada. Get your priorities straight, Sam.
OpenAI LLC is owned by Microsoft to a certain extend. Currently MS has invested $13 billion and MS will be entitled to 49% of profits until it earns a profit of $92 billion. Then MS shares will revert to OpenAI non profit. Another way to invest in OpenAI is to invest in companies like Nvidia because they make the GPUs that are used by OpenAI but we are a bit late to the party since Nvidia market cap has already increased due to this. But if you think esearch hard enough, I think you might be able to predict other winners.
How about building agents that can design and build a house for 1/5th the current insane price of homes in the USA and get 50% of the population out of their misery for housing or rent
And yet people in the United States have greater access to housing and food than any people have ever enjoyed anywhere in all history. Marxist dialogue from Marxist societal brainwashing.
It’s not about the construction costs, it’s about the animals treating it as monopoly and buying everything up for investment purposes and driving up prices. It’s fucked up.
Question i would ask @Sam Altman What could be the future equivalent in AI of Google’s victory over Yahoo, where Google’s page-rank algorithm outperformed Yahoo’s directory-based search approach
with increase in value for the company you are also decreasing the market share of the company by making it easy for competition to copy the same buisness model
Let me tell you something. If they say you have to worry, the models will get better and better - don't worry, they try to raise more money and build hype. Once you hear them saying don't worry, its all gonna be allright its nowhere close taking your jobs - then we are fk-ed
Open AI won't steamroller many others, and probably none when it comes to AGI as LLMs, however big they are, will never reach it on their own. Unless Open AI are secretly developing other models in parallel with LLMs they'll in fact be left behind....
large Language model read and create Unified golden record onto blockchain web3 as Json file. Ai agent user interface to interact with Json file on web3.
When I look forward I really hope that Sam Altman loses his annoying vocal fry. Listening to him talk ten or more years ago his voice was fairly normal but now that he's adopted vocal fry I find it extremely irritating to listen to anything he says.
First sentence from the CEO of OpenAI: 0:51 “Thanks for ‘dinnay’” Sam is human, confirmed. Almost thought he said “thanks for dinner”, but he definitely did not 😂
Sam Altman you're such a liar I'm going to go to every video that I see your face and tell everybody about how you steal people's data and say you created something
Subscribe to the 20VC UA-cam channel for more great interviews: www.youtube.com/@20VC
Choosing a concept space to stake your next software development project on or create a startup around these days feels like trying to build a sand castle on a beach being hit by a tsunami.
Yes and no. I think the principles he described ("don't try and plug gaps in models, but build specialized products that get better as models get better" are relatively straightforward to follow.
This is true. Yes OpenAI is a model company. You have to be far away from AI to win. But also have to have critical speed because OAI is working outward to every other space. Say you are a food delivery service and you want to use AI. It can help you until there is a consolidation of food delivery that uses the best model that improves cost, speed and quality. To the sand analogy. Build it near the boardwalk and hope the water doesn’t get close.
He says, "Don't build a patch for the current models" and you'll be okay. I'd like to know if ChatGPT Search and Custom GPTs are "patches". The truth is that anything that their models can do, they reserve the right to do for themselves.
@@motobanite No they are not at all straightforward because what constitutes a "patch" is completely open to interpretation.
The more I listen, the more this sounds like nonsense. "Bet on the models getting better." We are talking about ARTIFICIAL INTELLIGENCE. Intelligence literally spans everything. So there is no limit to what "better" includes. A model that can drive a car, run a country, colonize Mars.... There is nothing out of scope for a model that simulates human intelligence.
Master of speaking without saying anything
you have to be smart enough to read between the lines
Glad to hear the possibility of no-code tools empowering non-technical founders. I am building Licode, a no-code platform for building AI apps. Excited to see where it can go. :)
It could easily make that startup obsolete as well. But i guess we could always use the model itself to innovate and create new things. Im building a language with an interesting idea with the help of AI. Im moving with super human speed and i don't have any coding experience.
In the bin unless you are from a rich family with a history of committing sexual assault. Face facts. There hasn’t been anyone in this space who didn’t already start out rich for 20 years. Time to forget the startup lie and build and sell a product that works to someone that needs it.
Sad but true: “There's a lot of people in the world that I'm sure would be phenomenal AI researchers had their life paths just gone a little bit differently”
Yes, and everyone in the world would be at least past the poverty line had their life paths just gone a little bit differently too.
This was super interesting and Sam is *very* good at interviews now.
His programming has improved with iteration
🇧🇷🇧🇷🇧🇷🇧🇷👏🏻, Sam Altman is an exceptionally gifted leader with a sharp, visionary perspective that few possess. His keen sense of direction and strategic insight set him apart as a truly unique visionary. He's gifted that way!
Absolutely
@@XShollaj
He's a power obsessed maniac, look at what he's doing with world coin
Absolutely incredible conversation. What a time to be alive!
Interesting interview. A bit over obsessed with "being in trouble for this one" - the point of the interview is to ask probing interesting questions, it was funny the first time but got old!
Thanks Harry and Sam! Great episode.
Thanks for dinîayh 😊
Sammy would surely never lie to us
He never did it in the past either! :p
Sure
@@RusuTraianCristian yeah like naming company Open.. saying he doesn't have equity there (while he has options which is same thing).. this guy is sociopath
Great interview, Harry!
Sam seems very smart, though his responses feel a bit like incoherent based assumptions rather than clear strategic reasoning, but maybe that’s just my impression.
he’s a little bit frantic, but I think his mind is just moving a 1000 MPH; not a bad thing in my opinion but I see what you are saying.
6:06 Rather Build something to go Deliver the great AI tutor or the great AI medical advisor. 32:44 AI Lawyer, AI CAD Engineer
16:44 one of the Retirement thing I fantasies about doing is writing a book of everything I have learnt of how to build an organization & a culture that does this thing ( Research Lab ) not an organization that copies what everybody else has done.
23:39 Anthropic model for coding?
34:35 Cursor
33:58 for an expert example of how to question dodge
I like the last question, I tend to think the same as Sam, it's one thing to improve the tech but it's another to change the world. Especially the government, I think it will take time for these technologies to be widespread. I mean it's one thing to have CoPilot helping you to write an email. Or using UiPath RPA to automate some tasks. But it's another to replace entire jobs/teams with AIs.
first one here , probably will love this convo as always
Completely aligned with Human tendency to overestimate Speed and underestimate Scope of Societal Change enforced by Disruptive Technologies. Valid in the start of the Technosphere and Digisphere, valid in the start of the Cognosphere.
Heyooo love the podcasts of 20VC
CRAZY STUFF 🎉🚀
01 is definitely dropping tomorrow given the timing of this
well nope
Excellent interview. Sam shared some helpful insights. there’s tons of opportunity to apply powerful AI agents to benefit the world in many verticals. For example, ensuring people aren’t exposed to toxins from plastics (such as BPAs from food storage, etc.) and many similar products that we all use everyday. There are many of these kinds of issues, and an AI can probably help identify the most important ones to work on first.
Sam's vocal fry is through the roof
Great interview! Thank you
26:39 the heaviest thing in life is node_modules, Harry. It's node_modules. We should move on from JavaScript instead of pushing AGI agenda , agentic worlds, yada yada. Get your priorities straight, Sam.
Does he have stocks we can invest in? And what are they on the stock exchange?
OpenAI LLC is owned by Microsoft to a certain extend. Currently MS has invested $13 billion and MS will be entitled to 49% of profits until it earns a profit of $92 billion. Then MS shares will revert to OpenAI non profit. Another way to invest in OpenAI is to invest in companies like Nvidia because they make the GPUs that are used by OpenAI but we are a bit late to the party since Nvidia market cap has already increased due to this. But if you think
esearch hard enough, I think you might be able to predict other winners.
@@jean-phil100%
Am I the only one who’s more impressed with Sam’s PR game then his actual company
Is this actually sam altman talking or an ai?
You made me laugh
“There's a lot of people in the world that I'm sure would be phenomenal AI researchers had their life paths just gone a little bit differently”
How about building agents that can design and build a house for 1/5th the current insane price of homes in the USA and get 50% of the population out of their misery for housing or rent
And yet people in the United States have greater access to housing and food than any people have ever enjoyed anywhere in all history.
Marxist dialogue from Marxist societal brainwashing.
It’s not about the construction costs, it’s about the animals treating it as monopoly and buying everything up for investment purposes and driving up prices. It’s fucked up.
Generally speaking, the cost of homes is nothing in comparison to the cost of land.
@@loukah4401facts tho
"i would encourage people to be aligned with that".
That's not ominous at all.
the facilitator is a bit ego centric
Amazing!!
Question i would ask @Sam Altman
What could be the future equivalent in AI of Google’s victory over Yahoo, where Google’s page-rank algorithm outperformed Yahoo’s directory-based search approach
with increase in value for the company you are also decreasing the market share of the company by making it easy for competition to copy the same buisness model
Really great questions.
Damn Harry’s interview skills are getting strong! He could be the next TF
Strategic shoutout to Cursor considering OpenAi invested in it
Let me tell you something. If they say you have to worry, the models will get better and better - don't worry, they try to raise more money and build hype. Once you hear them saying don't worry, its all gonna be allright its nowhere close taking your jobs - then we are fk-ed
Causal reasoning is coming next
Open AI won't steamroller many others, and probably none when it comes to AGI as LLMs, however big they are, will never reach it on their own. Unless Open AI are secretly developing other models in parallel with LLMs they'll in fact be left behind....
OpenAI's o1 model is not just a better-reasoning LLM; it represents a paradigm shift in how foundation models are developed and deployed.
@@jean-phil In that case the full o1 must be much better than the preview version cos that hasn't a clue when it comes to human level reasoning
Congrats
the host needs to invest in a pants startup for sure!
This might be a hot take, but Sam is a warm person to me. I'm naturally trusting of his vibe despite what people say.
large Language model read and create Unified golden record onto blockchain web3 as Json file.
Ai agent user interface to interact with Json file on web3.
Guys we can look after our brains we need help in livingg❤❤🎉🎉senses
what is up with the amount of ads on this video lol, insane
ever heard of ad blockers ?
he just said dont develop any app as open ai your backend
imagine you are one of the Engineers and have to listen to this level of BS daily at work
Even for $50 000 a month it will be a real pain in the ear
this was already leaked
When I look forward I really hope that Sam Altman loses his annoying vocal fry. Listening to him talk ten or more years ago his voice was fairly normal but now that he's adopted vocal fry I find it extremely irritating to listen to anything he says.
FFFIIIIIRRRSTTT!!
First sentence from the CEO of OpenAI:
0:51 “Thanks for ‘dinnay’”
Sam is human, confirmed.
Almost thought he said “thanks for dinner”, but he definitely did not 😂
I think "thanks for having me today" and "thanks for doing it" got mixed up. It happens 😆
Ageism
Good god, take some of the $1B you raised and hire a communications coach. Stop talking like a frog.
Watched this leaked
Lol at could. Everyone could predict everything 😅
Trousers?
Second
Sam Altman you're such a liar I'm going to go to every video that I see your face and tell everybody about how you steal people's data and say you created something
Can you elaborate?
Nobestudy intelligence is my choice
Cursor 🫶