I saw this when it was uploaded on another channel. My main takeaway is that AI in the short run will help mathematicians develop proofs, but in the long run probably replace humans entirely.
Timestamps (Powered by Merlin AI) 00:09 - Discussion featuring renowned mathematicians and AI experts on future developments. 02:29 - Exploring innovative mathematics and reasoning methods for enhanced collaboration and discovery. 06:28 - Mathematics collaboration can now leverage AI and diverse skill sets. 08:30 - Collaborative approaches in mathematics are evolving with formal verification methods. 12:27 - Humans have a better intuition for math, but AI can assist in verification and generating counterexamples. 14:28 - Iterative techniques help in mathematical proofs but have limitations in complexity. 18:34 - AI will enhance but not replace traditional mathematical reasoning. 20:31 - Emerging abstractions in math could reshape education and research methodologies. 24:27 - AI will transform math learning by enhancing efficiency and understanding. 26:14 - Human understanding and aesthetics in mathematics are crucial for future developments. 29:52 - Math's formal verification could revolutionize scientific progress through automation. 31:50 - Mathematics will evolve with AI collaboration while preserving traditional practices. 35:35 - The impact of AI on accessibility and collaboration in mathematics. 37:21 - AI could democratize math, enhancing accessibility for scientific applications. 41:06 - Math now requires broad collaboration and AI proficiency for future success. 42:56 - AI tools will significantly enhance efficiency in mathematics. 46:46 - Integrating AI models for diverse reasoning types can revolutionize problem-solving. 48:29 - AI complements human problem-solving in math but struggles with data-scarce reasoning. 52:19 - Collaboration and infrastructure are key for advancing mathematics technologies. 54:14 - The future of mathematics includes creating a unified database of theorems. 58:07 - Mathematics reveals surprises that challenge current understanding and model limitations. 1:00:15 - Future AI models will improve precision and reference sourcing for drug discovery. 1:04:00 - AI models will increasingly handle complex reasoning tasks autonomously. 1:06:03 - AI is reshaping the relationship between mathematics and applied sciences. 1:10:13 - Integrating AI models can enhance reasoning and creativity in mathematical proofs. 1:12:09 - Diverse approaches can enhance problem-solving in AI, despite challenges in management. 1:16:29 - Empirical studies are crucial for advancing AI architectures in mathematical reasoning. 1:18:25 - Large collaborative projects will shape the future of math. 1:22:23 - AI can transform science and math, improving regulatory processes. 1:24:30 - Community engagement and networking opportunities are being enhanced.
i will always listen to terry tao talk. He is a powerhouse in his field and will be remembered for a very long time - both for his contributions to mathematical science, and his outreach
Listening to this conversation, it's incredible how rapidly the landscape of mathematics and science is evolving. Just a few years ago, the idea of machine assistants not only crunching numbers but also generating new ideas and verifying reasoning seemed like science fiction. Now, it's becoming a reality that just needs fine-tuning and adaptation. Of course, there are still challenging questions: How can we ensure the accuracy of AI's suggestions? How can we interact with AI in a way that truly broadens our intellectual horizons instead of reducing mathematics to a mechanical routine? The overall sentiment is clear: we're on the brink of a new era where mathematics, science, and AI are deeply interconnected, paving the way for a deeper and faster understanding of the world around us. In fact, we've already stepped into this new era.
Qué bueno sería que simultáneamente tuvierais una versión traducida al español en audio, en algún soporte vuestro, sé que RUclips de momento no permite esta opción, pero los millones de hispanos que os seguimos estaríamos encantados de escuchar la versión traducida. Muchas gracias ❤
I suggested once "new number shapes colors" - basically new base material understanding and then add "world knowledge" - and maybe like 5 senses sciences like cognitive science but also adaptive to "hyper" things - it's sort of touch and go if I understand the idea in relation to the world correctly - but it seems so fascinating - it just seems like there are so many of these strange precise innovative expansions of what the mind once believed was the limit of possibility
Don’t forget, all it takes to compete with trillions of dollars of investment and innovation to change the world is a Library card. You might even achieve feats with impact that of these gentlemen who are changing technology right before our eyes.
I am a mathematics professor and an independent researcher. If I manage, with the help of ChatGPT, to reach a new result in mathematics, such as proving the Riemann Hypothesis, will I be able to officially document this result as attributed to me? Additionally, I would like to know if OpenAI has a specialized mathematics committee to verify or refute the validity of this result.
Do you really think that the most important thing is whether it will attributed to you? In my opinion, the most important thing is whether the result will be verifiable by humans.
@Tommy_007 sure it should be verifiable by human, by the usage of AI, chatGPT is just for minimize the time to achieve the goal. In addition I think it must another way to confirm new proposal in mathematics other that used in universities system
@@Tommy_007 His question does really make sense. The world where we live is based on competition. In the field of mathematics, we all know Fields medal and Abel prize, and these strongly depend on our achievements, which depend on our talent and hard work Once we start by getting aid from AI, that means that the problem-solving skills and resilience that till today were necessary requirements will no longer have this role. It follows that your achievement is AI-dependent and, as such, you can no longer be awarded for something that's driven by AI. If we accept and welcome AI - as i believe it's the case for each of us - we also must accept we no longer play the same role as before. We must accept that our achievement is entirely determined by AI. Why should i award someone who, all of sudden, can find new opportunities across billions of data and solve incredibly complex problems that only few great mathematicians (Terence Tao is one of them) would be able to solve, thanks to the AI? I think there's nothing to be proud of.
Mark, I think Pro needs ( direct and convenient ) access to the full codebase, and to run code in an emulator to exploit the better reasoning. I paid for a month of Pro and I want to get value out of it but it is a bit disappointing in its current version. I am wasting that compute with such a limited context for the model to reason with. I feel like there is a bit of miss with Canvas, it seems like it wants to be just a bit more than it is. Make Pro have a better version of the current Canvas and you’ll sell a lot more subscriptions. This is my anecdotal experience, I am not a 10x programmer ( obviously ) but people like me are your target market.
Why is the model in the o1 system card an earlier checkpoint than the o1 and o1 pro that was released? The models that have been released are not the same ones that have been compared to benchmarks in the system card (if that even matters at this point). This goes completely against the entire purpose of releasing system cards and your preparedness framework. The capabilities of these systems are not accurately represented and evaluated. This is unacceptable behavior from a frontier AI company.
🤔 AI can really unlock the potential of citizen scientists. If there was some interface where regular people can input their ideas towards a given problem, perhaps AI can gather all that data and filter out novel ideas. These novel ideas could then be suggested to real scientists to determine if a real breakthrough has occurred.
No matter how hard you try to advertise this PRO version, I guarantee that most people wouldn’t subscribe to it, even if it cost $50 a month. Most people won’t even be convinced to go for the PLUS version.
Mark talked about teaching the models to do mathematical reasoning and needing examples to show models what incorrect reasoning looks like. I just want to say, there is no such thing as incorrect mathematical reasoning, if you tell me 2+2 is 5 idk what you did but you didn’t use mathematical reasoning, had you done so you would have gotten the right answer.
For a company is not expensive and they might use it so that it can pay off, for individual is expensive yes but also individuals have a very few percentage who would actually use it for what they pay for, I don't think that everyone who is not a business is using the 200$ on max capacity, and by that I mean every day, for the sole purpose to make money or for serious research in any field... There is a huge percentage who are using as entertainment and for making their life easyer, and thats all, for helping out and entertainment, the free and the 20$ subscription is plenty.
The title makes it seem like o1's related to the math stuff, but that's a total lie. They only talk about math provers and future AI; o1's completely irrelevant.
Do you think maths and AI can tackle ethical non-naturalism moral properties that are fundamental and cannot be reduced to natural properties? Goodness, rightness, wrongness, obligation, duty, respect
@@GodbornNoven Scenario: A surgeon has five patients in need of organ transplants to survive. A healthy individual, with compatible organs, visits for a routine check-up. The surgeon considers sacrificing this person to harvest organs and save the five patients. Analysis: Most people find the idea of killing one person to save five morally abhorrent, despite the numerical advantage. This intuition supports the notion that certain actions are intrinsically wrong, regardless of outcomes.
Scenario: Imagine a machine capable of providing you with any desired experience, indistinguishable from reality. Once connected, you would live a life filled with pleasure and satisfaction, unaware that these experiences are artificially generated. You have the option to plug into this machine for the rest of your life or continue living your actual life, with all its inherent uncertainties and challenges. Analysis: Many individuals intuitively reject the idea of connecting to the machine, valuing authentic experiences over artificial pleasure. This suggests that certain moral values, such as authenticity, possess an intrinsic worth that cannot be reduced to mere pleasure or natural properties.
Je trouve l’IA passionnante, mais elle me fait également un peu peur, notamment en raison de l’idée que nous pourrions nous-même être des consciences artificielles. Si nous arrivons un jour à créer des consciences artificielles, rien n’exclut que nous soyons déjà, nous aussi, des consciences “simulées”. Par ailleurs, des phénomènes mystérieux de la physique, comme les constantes universelles ou le comportement étrange des particules quantiques, pourraient tout à fait correspondre à des « paramètres » prédéfinis dans un monde artificiel. À l’heure actuelle, nous n’avons pas d’explication satisfaisante pour ces mystères ; ils paraissent presque illogiques, voire aberrants, ce qui renforce l’idée qu’ils pourraient résulter d’une réalité non “naturelle” mais simulée.
$200 is not for richest by far pretty much in any country at this point. Name a country that is so poor that you can hire a mathematician for less than $200 per month that will work as much as you like.
AGI is a very slippery concept. One could argue that GPT-4O is AGI. I’m quite a practical person, and I find OpenAI’s definition quite powerful: AGI will be able to perform most economically valuable human tasks. Let’s see about that. GPT-4 was quite impressive, but it makes so many mistakes that you really can’t let it work on its own. We’ll see what happens when fine-tuning and agentic tools start using GPT-01. it might become accurate enough to handle some tasks independently. Previous models have already saved Amazon millions on simple and repetitive coding tasks. I’m pretty sure this new “wave” will unlock many more tasks for automation. But frankly, I have zero expectations that GPT-01 will be able to perform most economically valuable tasks on its own.
"hello OpenAI i would like to present an idea that might useful for your development and to increase collaboratio between intelligent organization The idea to integrate artificial intelligence white other organizations working in the same field I believe that this collaboratio will help accalarate the progress and development artificiel intelligence and will incrasse interation between developers and users I hope you find this idea appealing and that it "becomes a step towards futher futur achievementsThan k you for your attention and time
You’re making that connection based on a headline. It’s disgusting for a newspaper to use a suicide to take a shot at a company. It’s pathetic that you are doing the same thing.
@@InternetStranger10101 How convenient that whistleblowers always kill themselves while they're building a case against a major corporation. But just a coincidence XD
Nobody in their right mind would pay 200 dollars a month for unlimited access to the model, if you want someone to spend that kind of money on this, make it a permanent purchase where people can self-host the service, so you don't have to host it on your own servers.
Suchir Balaji's death is very sad, such a bright human. There is definitely some accountability that openAI an it's competitors need to own for the cultural impact of their insatiable desire for progres.
I saw this when it was uploaded on another channel. My main takeaway is that AI in the short run will help mathematicians develop proofs, but in the long run probably replace humans entirely.
Wow thank you for the summary bro. Saved me an hr
Just use video AI summarizer lol
Timestamps (Powered by Merlin AI)
00:09 - Discussion featuring renowned mathematicians and AI experts on future developments.
02:29 - Exploring innovative mathematics and reasoning methods for enhanced collaboration and discovery.
06:28 - Mathematics collaboration can now leverage AI and diverse skill sets.
08:30 - Collaborative approaches in mathematics are evolving with formal verification methods.
12:27 - Humans have a better intuition for math, but AI can assist in verification and generating counterexamples.
14:28 - Iterative techniques help in mathematical proofs but have limitations in complexity.
18:34 - AI will enhance but not replace traditional mathematical reasoning.
20:31 - Emerging abstractions in math could reshape education and research methodologies.
24:27 - AI will transform math learning by enhancing efficiency and understanding.
26:14 - Human understanding and aesthetics in mathematics are crucial for future developments.
29:52 - Math's formal verification could revolutionize scientific progress through automation.
31:50 - Mathematics will evolve with AI collaboration while preserving traditional practices.
35:35 - The impact of AI on accessibility and collaboration in mathematics.
37:21 - AI could democratize math, enhancing accessibility for scientific applications.
41:06 - Math now requires broad collaboration and AI proficiency for future success.
42:56 - AI tools will significantly enhance efficiency in mathematics.
46:46 - Integrating AI models for diverse reasoning types can revolutionize problem-solving.
48:29 - AI complements human problem-solving in math but struggles with data-scarce reasoning.
52:19 - Collaboration and infrastructure are key for advancing mathematics technologies.
54:14 - The future of mathematics includes creating a unified database of theorems.
58:07 - Mathematics reveals surprises that challenge current understanding and model limitations.
1:00:15 - Future AI models will improve precision and reference sourcing for drug discovery.
1:04:00 - AI models will increasingly handle complex reasoning tasks autonomously.
1:06:03 - AI is reshaping the relationship between mathematics and applied sciences.
1:10:13 - Integrating AI models can enhance reasoning and creativity in mathematical proofs.
1:12:09 - Diverse approaches can enhance problem-solving in AI, despite challenges in management.
1:16:29 - Empirical studies are crucial for advancing AI architectures in mathematical reasoning.
1:18:25 - Large collaborative projects will shape the future of math.
1:22:23 - AI can transform science and math, improving regulatory processes.
1:24:30 - Community engagement and networking opportunities are being enhanced.
Thanks Merlin
I hope google comes with an AI to automate this important buy difficult talk of timestamping.
i will always listen to terry tao talk. He is a powerhouse in his field and will be remembered for a very long time - both for his contributions to mathematical science, and his outreach
It’s always a pleasure to listen to the great Terence Tao
Listening to this conversation, it's incredible how rapidly the landscape of mathematics and science is evolving. Just a few years ago, the idea of machine assistants not only crunching numbers but also generating new ideas and verifying reasoning seemed like science fiction. Now, it's becoming a reality that just needs fine-tuning and adaptation.
Of course, there are still challenging questions: How can we ensure the accuracy of AI's suggestions? How can we interact with AI in a way that truly broadens our intellectual horizons instead of reducing mathematics to a mechanical routine?
The overall sentiment is clear: we're on the brink of a new era where mathematics, science, and AI are deeply interconnected, paving the way for a deeper and faster understanding of the world around us. In fact, we've already stepped into this new era.
Qué bueno sería que simultáneamente tuvierais una versión traducida al español en audio, en algún soporte vuestro, sé que RUclips de momento no permite esta opción, pero los millones de hispanos que os seguimos estaríamos encantados de escuchar la versión traducida. Muchas gracias ❤
use ElevenLabs to translate the whole video. works well!
Hopefully these tools will help us do better at getting more kids interested in mathematics, to see the beauty and utility.
35 minutes in, and this is wonderful. Thank you.
I didn’t expect this crossover
I love math. I hope AGI of OpenAI can forward math overall. Can't wait to see the results :)
It would be great to have this video available for any languages.
Thank you for publishing this ☺
The future of maths: AI does the maths.
I suggested once "new number shapes colors" - basically new base material understanding and then add "world knowledge" - and maybe like 5 senses sciences like cognitive science but also adaptive to "hyper" things - it's sort of touch and go if I understand the idea in relation to the world correctly - but it seems so fascinating - it just seems like there are so many of these strange precise innovative expansions of what the mind once believed was the limit of possibility
Don’t forget, all it takes to compete with trillions of dollars of investment and innovation to change the world is a Library card. You might even achieve feats with impact that of these gentlemen who are changing technology right before our eyes.
I am a mathematics professor and an independent researcher. If I manage, with the help of ChatGPT, to reach a new result in mathematics, such as proving the Riemann Hypothesis, will I be able to officially document this result as attributed to me?
Additionally, I would like to know if OpenAI has a specialized mathematics committee to verify or refute the validity of this result.
Just write it down and say you wrote it yourself. That's how i do my tests hahahaha
Do you really think that the most important thing is whether it will attributed to you? In my opinion, the most important thing is whether the result will be verifiable by humans.
@Tommy_007 sure it should be verifiable by human, by the usage of AI, chatGPT is just for minimize the time to achieve the goal. In addition I think it must another way to confirm new proposal in mathematics other that used in universities system
@@Tommy_007 His question does really make sense.
The world where we live is based on competition. In the field of mathematics, we all know Fields medal and Abel prize, and these strongly depend on our achievements, which depend on our talent and hard work
Once we start by getting aid from AI, that means that the problem-solving skills and resilience that till today were necessary requirements will no longer have this role.
It follows that your achievement is AI-dependent and, as such, you can no longer be awarded for something that's driven by AI.
If we accept and welcome AI - as i believe it's the case for each of us - we also must accept we no longer play the same role as before. We must accept that our achievement is entirely determined by AI.
Why should i award someone who, all of sudden, can find new opportunities across billions of data and solve incredibly complex problems that only few great mathematicians (Terence Tao is one of them) would be able to solve, thanks to the AI?
I think there's nothing to be proud of.
Hello, please answer this question. With the $20 "ChatGPT o1" plan, how many math-related questions can be asked daily?
Wonderful conversation. Thank you.
At 52:54 Terence started using 0.002% of his brain power
I already saw this on another channel. OpenAi thinks its important enough to post on main
Mark, I think Pro needs ( direct and convenient ) access to the full codebase, and to run code in an emulator to exploit the better reasoning. I paid for a month of Pro and I want to get value out of it but it is a bit disappointing in its current version. I am wasting that compute with such a limited context for the model to reason with. I feel like there is a bit of miss with Canvas, it seems like it wants to be just a bit more than it is. Make Pro have a better version of the current Canvas and you’ll sell a lot more subscriptions. This is my anecdotal experience, I am not a 10x programmer ( obviously ) but people like me are your target market.
"Would you like to create a Hollywood action movie or an animated film using generative AI, like in Sora?"
timestamp
in sora is it possible to use image with a human in it, with the 200us dollar pro version
Why is the model in the o1 system card an earlier checkpoint than the o1 and o1 pro that was released? The models that have been released are not the same ones that have been compared to benchmarks in the system card (if that even matters at this point). This goes completely against the entire purpose of releasing system cards and your preparedness framework. The capabilities of these systems are not accurately represented and evaluated. This is unacceptable behavior from a frontier AI company.
Thank you!
🤔 AI can really unlock the potential of citizen scientists. If there was some interface where regular people can input their ideas towards a given problem, perhaps AI can gather all that data and filter out novel ideas. These novel ideas could then be suggested to real scientists to determine if a real breakthrough has occurred.
Mark is a little too naively optimistic here in my perspective.
🇧🇷🇧🇷🇧🇷🇧🇷👏🏻, Wow! Amazing.
Music is all calculable, harmony, rhythm, and melody, in that order.
No matter how hard you try to advertise this PRO version, I guarantee that most people wouldn’t subscribe to it, even if it cost $50 a month. Most people won’t even be convinced to go for the PLUS version.
trence tao 교수님 더더더드드듬기 스킬 장난아니네
Awesome!
Mark talked about teaching the models to do mathematical reasoning and needing examples to show models what incorrect reasoning looks like. I just want to say, there is no such thing as incorrect mathematical reasoning, if you tell me 2+2 is 5 idk what you did but you didn’t use mathematical reasoning, had you done so you would have gotten the right answer.
Mathematics - Physics - Chemistry - Engineering - Biology - Neuroscience - Psychology - Statistics - Sociology - Economics... ⤴👁
😎🤖
Remember Suchir Balaji.
What is Math?
What is math?
Throwing this video in GPT o1 to give me a summary will be wonderful.
o1 is not gpt
For a company is not expensive and they might use it so that it can pay off, for individual is expensive yes but also individuals have a very few percentage who would actually use it for what they pay for, I don't think that everyone who is not a business is using the 200$ on max capacity, and by that I mean every day, for the sole purpose to make money or for serious research in any field... There is a huge percentage who are using as entertainment and for making their life easyer, and thats all, for helping out and entertainment, the free and the 20$ subscription is plenty.
Being here just bcs of Terence Tao. 🤭
More credit
Pelis buat sistem memori Chat gpt permanen pelis
When no one has a job anymore then who pays the AI?
The title makes it seem like o1's related to the math stuff, but that's a total lie. They only talk about math provers and future AI; o1's completely irrelevant.
Too late, o1 just wrote o2!
Do you think maths and AI can tackle ethical non-naturalism moral properties that are fundamental and cannot be reduced to natural properties? Goodness, rightness, wrongness, obligation, duty, respect
No, of course not.
Does that matter
@@GodbornNoven Scenario: A surgeon has five patients in need of organ transplants to survive. A healthy individual, with compatible organs, visits for a routine check-up. The surgeon considers sacrificing this person to harvest organs and save the five patients.
Analysis: Most people find the idea of killing one person to save five morally abhorrent, despite the numerical advantage. This intuition supports the notion that certain actions are intrinsically wrong, regardless of outcomes.
Scenario: Imagine a machine capable of providing you with any desired experience, indistinguishable from reality. Once connected, you would live a life filled with pleasure and satisfaction, unaware that these experiences are artificially generated. You have the option to plug into this machine for the rest of your life or continue living your actual life, with all its inherent uncertainties and challenges.
Analysis: Many individuals intuitively reject the idea of connecting to the machine, valuing authentic experiences over artificial pleasure. This suggests that certain moral values, such as authenticity, possess an intrinsic worth that cannot be reduced to mere pleasure or natural properties.
Je trouve l’IA passionnante, mais elle me fait également un peu peur, notamment en raison de l’idée que nous pourrions nous-même être des consciences artificielles. Si nous arrivons un jour à créer des consciences artificielles, rien n’exclut que nous soyons déjà, nous aussi, des consciences “simulées”.
Par ailleurs, des phénomènes mystérieux de la physique, comme les constantes universelles ou le comportement étrange des particules quantiques, pourraient tout à fait correspondre à des « paramètres » prédéfinis dans un monde artificiel. À l’heure actuelle, nous n’avons pas d’explication satisfaisante pour ces mystères ; ils paraissent presque illogiques, voire aberrants, ce qui renforce l’idée qu’ils pourraient résulter d’une réalité non “naturelle” mais simulée.
Elon musk bought Gene Wilders house because he's a kingpin at mixing poo and chocolate together they call him the master chocolateer
AI
⬇️
⬇️
System 01 : Chat GPT-1, GPT-2, GPT-3, GPT-4.
System 02 : model o1, o2, → oNth (possibly).
⬇️
⬇️
⬇️
AGI . . .
A $200 plan can't benefit all of humanity, only the richest who can affod it.
They are trying to fund their pursuit of AGI
Devin: hold my beer
crazy right?? turns out computers require money to run, who'da thunk it????
$200 is not for richest by far pretty much in any country at this point. Name a country that is so poor that you can hire a mathematician for less than $200 per month that will work as much as you like.
TBF, they said AGI, not o1.
It's discussion not maths solving example on chatgpt useless discussion
all of humanity with $200 a month? Better to just not say that, right?
6day old video
Ok and?
The rise of Skynet
Considering humanity has been at constant war amongst itself since the dawn of time in all groups, what you speak of is irrelevant to us.
MATRIX AND TERMINATOR COMING SOON YEARS AND END OF HUMANITY DESTROY/KILLING BY AI ...😬😬😬😱😱😱💀💀💀☠️☠️☠️🩸🩸🩸
ها
There is less than 7 years until AGI.
We already have "AGI". o1 is AGI.
There's less than two years left until ASI.
@@pandoraeeris7860 we don't have agi until an llm can reason like a person
AGI is a very slippery concept. One could argue that GPT-4O is AGI.
I’m quite a practical person, and I find OpenAI’s definition quite powerful: AGI will be able to perform most economically valuable human tasks.
Let’s see about that. GPT-4 was quite impressive, but it makes so many mistakes that you really can’t let it work on its own.
We’ll see what happens when fine-tuning and agentic tools start using GPT-01. it might become accurate enough to handle some tasks independently.
Previous models have already saved Amazon millions on simple and repetitive coding tasks. I’m pretty sure this new “wave” will unlock many more tasks for automation.
But frankly, I have zero expectations that GPT-01 will be able to perform most economically valuable tasks on its own.
@@pandoraeeris7860 bro, AGI needs to first do stuff an IQ100 person can do on a computer, *reliably*
Source: trust me bro
"hello OpenAI i would like to present an idea that might useful for your development and to increase collaboratio between intelligent organization The idea to integrate artificial intelligence white other organizations working in the same field I believe that this collaboratio will help accalarate the progress and development artificiel intelligence and will incrasse interation between developers and users I hope you find this idea appealing and that it "becomes a step towards futher futur achievementsThan k you for your attention and time
☺☺☺☺☺😃
FYI, a whistleblower just died.
People die all the time, life is brutal. This is irrelevant to the agenda of OpenAI.
source?
@nadella1991 Can't post links. His name is Suchir Balaji.
You’re making that connection based on a headline. It’s disgusting for a newspaper to use a suicide to take a shot at a company. It’s pathetic that you are doing the same thing.
@@InternetStranger10101 How convenient that whistleblowers always kill themselves while they're building a case against a major corporation. But just a coincidence XD
200$ all humanyty lol!)))🤬
google AIstudio 2KK context window for free! you'r welcome!
2nd
Just saw that Sam Altman donated 1 million to Trump. I feel most people in the AI world, from Musk to Altman, are now corrupted. It makes me very sad.
1st
OpenAI murdered Suchir Balaji.
Making baseless claims is very stupid!
wow! she is beautiful
Nobody in their right mind would pay 200 dollars a month for unlimited access to the model, if you want someone to spend that kind of money on this, make it a permanent purchase where people can self-host the service, so you don't have to host it on your own servers.
why are you posting this here?
i think they know how to price the product better than you do.
Pro tip, you don't need to tell them how many people will pay that much, because they already know exactly how many people are paying that much.
Now you are murdering whistleblowers?
Suchir Balaji's death is very sad, such a bright human.
There is definitely some accountability that openAI an it's competitors need to own for the cultural impact of their insatiable desire for progres.