Will the stateful API make things faster or cheaper? Or does it just simplify the API usage? I'd guess mostly the latter, but don't know.
OpenAI's plans according to Sam Altman
21–30 of 70 posts
Re: OpenAI's plans according to Sam Altman
#22Earlier quoted context omitted.
As I understand it, most people's objection to the calls for regulation are based on the fact that regulation raises the barriers to entry for competition. Regulatory compliance can be expensive and favors those with deep pockets. The more AI startups one can prevent from starting/growing now, the fewer competitors there will be in the future.
All of which is entirely reasonable and so is considering all of Samas possible motivations. But just declaring another parties intentions as you see fit is such an incredibly bad style in any discourse. When filling the gaps of your knowledge, there has to be a world in which Sama is in fact a good human being – in addition to the world where he is not – when that assumption does not logically conflict with anything…
Perhaps, but Altman turned an open non profit org to a for profit closed org. So why should we continue to take his arguments at face value?
Re: OpenAI's plans according to Sam Altman
#23Earlier quoted context omitted.
As I understand it, most people's objection to the calls for regulation are based on the fact that regulation raises the barriers to entry for competition. Regulatory compliance can be expensive and favors those with deep pockets. The more AI startups one can prevent from starting/growing now, the fewer competitors there will be in the future.
All of which is entirely reasonable and so is considering all of Samas possible motivations. But just declaring another parties intentions as you see fit is such an incredibly bad style in any discourse. When filling the gaps of your knowledge, there has to be a world in which Sama is in fact a good human being – in addition to the world where he is not – when that assumption does not logically conflict with anything…
Guessing the true motivations of the rich and powerful is an essential element of democracy, because they never reveal their intentions.
This is also why even countries with stricter libel laws than the U.S. make broad exceptions for public figures. If you can't speculate about motivations in public, you'll have a dictatorship soon.
Re: OpenAI's plans according to Sam Altman
#24Is there a reason this post is pointing to a capture of the article at archive.org rather than to the original Humanloop webpage?
It was removed:
https://humanloop.com/blog/openai-plans
"This content has been removed at the request of OpenAI."
Re: OpenAI's plans according to Sam Altman
#25Whether scaling laws hold or not, is up for debate. What isn't up for debate, is that: 1) Giant serverfarms are expensive 2) People want on premises/on machine solutions 3) LoRA Tuning of small models continues to excel 4) Thus specialized models continue to evolve at a fast pace 5) Open source foundation models on which tuning can be done are accelerating by the week 6) Performance doesn't matter once a model is "go…
1) Giant serverfarms are expensive, but Microsoft has money and savings on scale. This can compare favourably to running locally. 2) People use what LLMs software makers give them and don't care at all about on premise. Software makers want something reliable they can deploy at scale, and not end up debugging problems on a client 4GB machine with an iGPU and multitude of OSs. The actually deciding people want to dele…
A long time ago, in order to get computers to do anything useful, the hardware required huge rooms.
Today, a little cell phone is significantly more powerful than a room sized computer back in the days.
LLMs shouldn't be any different. Today, they require giant server farms to run. Tomorrow, they will run on little robots/cell phones.
A few things will allow this:
1. Chip makers like Apple, Qualcomm, AMD, Intel, Nvidia, will heavily emphasize AI performance in their future chips. Expect accelerators like the Neural Engine to get significantly bigger. In the future, I expect that most of the transistors in a SoC will be dedicated to AI acceleration - and not CPUs/GPUs.
2. Moore's law is not dead. We will continue to get more transistors in a given area. For example, TSMC has plans for 1nm which will likely have 4-5x more transistors per area than the 5nm Apple M2 or around 100 billion transistors. For comparison, the 4090 has 76 billion transistors. A 1nm Apple Silicon Max chip could theoretically have around 300 billion transistors, which makes it 4x more than a 4090. If GPT-4 requires 4x 4090 to run, then a single 1nm M Max might be able to do it in the future.
3. Larger models will optimize to require less resources.
4. Smaller models will become more capable.
These forces will converge and we will have local LLMs that are top of the line.
Re: OpenAI's plans according to Sam Altman
#26Earlier quoted context omitted.
As I understand it, most people's objection to the calls for regulation are based on the fact that regulation raises the barriers to entry for competition. Regulatory compliance can be expensive and favors those with deep pockets. The more AI startups one can prevent from starting/growing now, the fewer competitors there will be in the future.
All of which is entirely reasonable and so is considering all of Samas possible motivations. But just declaring another parties intentions as you see fit is such an incredibly bad style in any discourse. When filling the gaps of your knowledge, there has to be a world in which Sama is in fact a good human being – in addition to the world where he is not – when that assumption does not logically conflict with anything…
There's one obvious pink elephant in safety that makes it very difficult for me to perceive positive intentions. What is a bigger threat to "safety" from AI? You and I - society society at large, or an organization whose entire purpose for existence is to think up brand new ways to violently impose their will on the rest of the world - and have a long history of showing minimal to zero moral constraints in pursuit, and execution, of those ends? Let's not forget to also factor in near to zero laws meaningfully constraining their behavior, and a trillion dollar annual budget to cook up new forms of violence.
Yet all the talk about "safety" is about keeping the chatbot away from you and I, while completely ignoring the real threat. And of course the real threat will always be granted access to whatever AI system it demands, because of those magical words that can make the sun rise in the West and 2+2 sum to 5: national security. And the version they're using isn't going to be responding with "As an AI language learning model I cannot..."
Where's the logic in this?
Re: OpenAI's plans according to Sam Altman
#27I’ve seen parts of the Congress testimony of Sam Altman amd it felt like the guy was on a power trip with his regulation ideas. What he accomplished with OpenAI is for the history books, but wanting monopoly over AI development in the US makes him look like an up and coming villain.
It's almost as if he's changed his tune since having 10b$ invested in his company.
Changed? It's the same song as before they were famous.
> We’ve co-authored a paper that forecasts how malicious actors could misuse AI technology, and potential ways we can prevent and mitigate these threats. This paper is the outcome of almost a year of sustained work with our colleagues at the Future of Humanity Institute, the Centre for the Study of Existential Risk, the Center for a New American Security, the Electronic Frontier Foundation, and others.
- 2018, https://openai.com/research/preparing-for-malicious-uses-of-...
Re: OpenAI's plans according to Sam Altman
#28Earlier quoted context omitted.
1) Giant serverfarms are expensive, but Microsoft has money and savings on scale. This can compare favourably to running locally. 2) People use what LLMs software makers give them and don't care at all about on premise. Software makers want something reliable they can deploy at scale, and not end up debugging problems on a client 4GB machine with an iGPU and multitude of OSs. The actually deciding people want to dele…
A long time ago, in order to get computers to do anything useful, the hardware required huge rooms. Today, a little cell phone is significantly more powerful than a room sized computer back in the days. LLMs shouldn't be any different. Today, they require giant server farms to run. Tomorrow, they will run on little robots/cell phones. A few things will allow this: 1. Chip makers like Apple, Qualcomm, AMD, Intel, Nvid…
We started from thin clients connecting to room-sized mainframes made by giant corporations, and we ended at small cellphone-size devices mostly acting as thin clients connecting to datacenter-building-sized services run by giant corporations, even though the small devices could theoretically do much more than the original thin clients.
Being able to run locally won't change much, if the local LLMs are owned by Microsoft, Google and Apple. Or if the 'local' LLM is actually an API call to a web service because the programmer/company decided to delegate the issue and not deal with all the software deployment issues. Or because Microsoft decided to hoover up all the data, and programmers decided it's so much easier to just call the OS API and not care whether it's running locally or remotely.
Yeah, you'll have a local LLM that is pretty good. There'll always be a bit better LLM running remotely, because most people won't run giant servers and don't care to update often. How much does any of that matter, when the AI is completely controlled by GiantCorp?
Re: OpenAI's plans according to Sam Altman
#29Earlier quoted context omitted.
A long time ago, in order to get computers to do anything useful, the hardware required huge rooms. Today, a little cell phone is significantly more powerful than a room sized computer back in the days. LLMs shouldn't be any different. Today, they require giant server farms to run. Tomorrow, they will run on little robots/cell phones. A few things will allow this: 1. Chip makers like Apple, Qualcomm, AMD, Intel, Nvid…
We started from thin clients connecting to room-sized mainframes made by giant corporations, and we ended at small cellphone-size devices mostly acting as thin clients connecting to datacenter-building-sized services run by giant corporations, even though the small devices could theoretically do much more than the original thin clients. Being able to run locally won't change much, if the local LLMs are owned by Micro…
It will honestly depend on just how much better the cloud versions are relative to the local versions.
If the cloud versions are significantly better (like they are now) than local LLMs, then cloud will continue to be the way. If local LLMs reach 95% of what cloud versions can do, then I think local ones might win out because the cost will be smaller, it will be faster (in latency), and it will have more privacy.
In 3-4 years? I'm willing to bet that local LLMs will have a sizable market.
I basically expect the neural engine inside an Apple Silicon M7 to be 80% of the SoC - instead of 10% like it is today. We're going to be buying NPUs (Neural Processor Unit) with a CPU and GPU attached to it. Right now, we're buying a CPU with a GPU and NPU attached to it.
I don't expect giant corporations to control LLMs. I expect smaller companies to be able to compete and that techniques for training and deploying LLMs will eventually look similar to how software is built today - where anyone can train and deploy LLMs.
Re: OpenAI's plans according to Sam Altman
#30Whether scaling laws hold or not, is up for debate. What isn't up for debate, is that: 1) Giant serverfarms are expensive 2) People want on premises/on machine solutions 3) LoRA Tuning of small models continues to excel 4) Thus specialized models continue to evolve at a fast pace 5) Open source foundation models on which tuning can be done are accelerating by the week 6) Performance doesn't matter once a model is "go…
1) Giant serverfarms are expensive, but Microsoft has money and savings on scale. This can compare favourably to running locally. 2) People use what LLMs software makers give them and don't care at all about on premise. Software makers want something reliable they can deploy at scale, and not end up debugging problems on a client 4GB machine with an iGPU and multitude of OSs. The actually deciding people want to dele…
1) Did you do the math? Makes no intuitive sense to me.
2) did you do a survey? Because I see no reason to just assume no one wants or needs on-prem.
3) you're extrapolating the current state of affairs into the future with no logical justification for doing so.
4) This just relies on 1-3 and is also only speculation. I can easily see scenarious where big tech fail to pivot into AI properly and go down.