4 ms·

AGI is a technology or a feature, not a product. ChatGPT is a product. They need some more products to pay for one of the most expensive technologies ever (to n

AGI is a technology or a feature, not a product. ChatGPT is a product. They need some more products to pay for one of the most expensive technologies ever (to not be delivered yet).

leptons 2y ago
It's a shame that $1 trillion is being poured into AI so quickly, but fusion research has only seen a fraction of that over many decades.
- andsoitis 2y ago
  Isn’t that a reflection, to some (major?) extent, of the general sense of likelihood of breakthrough?
  - xmprt 2y ago
    I used to think like this but after seeing the amount of money invested into crypto companies which most average people could have quickly dismissed as irrelevant, I'm not sure VCs are a good judge of value.
    - 9rx 2y ago
      That's not an entirely fair take. Once upon a time it was believed that cryptocurrencies would be able to supplant payment processors (i.e. VISA, Master Card) We even had a rollout of crypto ATMs to utilize the belief in that network. That is a money printing business, so it wasn't unreasonable for VC money to chase it. Sure, those intimately familiar with the technology may have always known that it wouldn't work, but the average person absolutely did not. They only came to learn that it wouldn't work as the cryptocurrency ventures proved that it wouldn't work. The pivot to being some kind of tool to store wealth, or a way to trade cartoon monkeys, or whatever as a last ditch effort to save the companies that failed to turn it into payment processing technology was more obviously irrelevant, but that came much, much later. The VCs were desperately seeking an exit by that point.
  - j_maffe 2y ago
    No it's the perception of likelihood of breakthrough
  - caseyy 2y ago
    It’s why we have warnings to investors saying past performance is not an indicator of future returns.
  - plorg 2y ago
    Maybe it means investors think it will happen, maybe it means investors simply think it will be profitable. It could even be investors seeking a market that has juice long after ZIRP dried up.
  - rchaud 2y ago
    Capital gains on AI investment doesn't need a breakthrough. I'm sure pharma companies and others are building their own custom models to assist R&D, but those aren't the ones sinking the truly huge sums into consumer-facing LLMs.
  - leptons 2y ago
    We already know fusion exists and happens and has been happening for billions of years - it's well understood. It may not be simple to reproduce and sustain on earth, but there is a fairly clear path to get there. There has never been an AGI, and there's no clear path to get there. And no, LLMs are not bringing us any closer to a real AGI, no matter how much Sam Altman wants to try to redefine what "AGI" means so that he can claim he has it. The difference between AGI funding and fusion funding is the people hawking AGI are better liars and the people funding are far more gullible than the people trying to make fusion power a reality.
    - Nevermark 2y ago
      We already know fusion has been a grind. For decades. We already know that computation has grown exponentially in scale and cost, for decades. With breakthrough software tech showing up reliably as it harnesses greater computational power, even if unpredictably and sparsely. If we view AI across decades, it has reliably improved. Even through its “winters”. It just took time to hit thresholds of usefulness making relatively smooth progress look very random from a consumers point of view. As computational power grows, and current models become more efficient, the hardware overhang for the next major step is growing rapidly. That just hasn’t been the case for fusion.
      - leptons 2y ago
        >We already know fusion has been a grind. For decades. Let's also not forget that "AI" has been grinding on for at least half a century, with nothing even coming close to AGI. No, LLMs are not AGI and cannot be AGI. They may be a component of an eventual AGI in the future, but we're still far, far away from artificial general intelligence. LLMs are not intelligent, but they do fool a lot of people. With the paltry amount of money spent on fusion, it's no wonder it's been delayed for so long. Maybe try putting the $1 Trillion into it in the short time "AI" has had its windfall, and let's see what happens with fusion. >If we view AI across decades, it has reliably improved That's your opinion. It's still not much better than Eliza. And LLMs lie half the time, just to make the user happy hearing what they want to hear. >As computational power grows, and current models become more efficient, the hardware overhang for the next major step is growing rapidly. And it's still going to "hallucinate" and give wrong answers a lot of the time because the fundamental tech behind LLMs is a bit flawed and is in no way even close to AGI. >That just hasn’t been the case for fusion. And yet we have actual fusion reactions happening now, sustained for 22 minutes. We're practically on the threshold of having fusion power, but still there is nowhere near the money being spent on it compared to LLMs, which won't be capable of AGI in spite of the marketing line you've been told.
- imafish 2y ago
  Unfortunately VC investments rely on FOMO, and there is no current huge breakthrough in fusion research that they are afraid of missing out on.
- Tenoke 2y ago
  Much, much less was being poured into AI until it started to have some returns.
  - leptons 2y ago
    What returns? OpenAI is still not profitable.
- csharpminor 2y ago
  If you believe the AGI thesis, this is a stepping stone to unlocking fusion and other scientific advances.
- immibis 2y ago
  Once we unlock fusion, humanity starts down a millenia-long path to making the planet permanently uninhabitable (not just temporarily like with climate change) by turning all the water into bitcoins.
  - orbifold 2y ago
    I don’t think that is true, there is only so much heavy water on the planet and doing hydrogen fusion is not all that economical.
  - leptons 2y ago
    "There are 1,335,000,000,000,000,000,000 liters of water in all the oceans." "A large fusion power plant, generating 1,500 megawatts of electricity, would consume approximately 600 grams of tritium and 400 grams of deuterium each day. A 1,000 MW coal-fired power plant requires 2.7 million tonnes of coal per year, while a similar fusion plant would need only 250 kg of fuel per year (half deuterium, half tritium)." You do the math. I'm personally not worried about running out of water due to fusion.
- bookofjoe 2y ago
  Gemini 10.0 Wait for it
- BeFlatXIII 2y ago
  It'll be fun to point and laugh at everyone when the bubble bursts.
  - ActorNightly 2y ago
    The bubble won't burst. LLMs are very useful. There just hasn't been enough optimization for general use case, which is kinda how all tech starts. Eventually someone will make a more optimized lower parameter model that has decent accuracy, and can be run on lower grade hardware, that will be a pretty good assistant with manual wrappers around automation.
    - leptons 2y ago
      This is the same llm-splaining that practically every enthusiast does. It's either "more hardware" or "more optimization" will solve all the problems and bring us into a new era of blah blah blah. But it won't. It won't solve LLM slop or hallucinations, which are inherent to how LLMs work, and why that bubble is already looking pretty unstable.
      - ActorNightly 2y ago
        The goal isn't to build perfect assistants with LLM. The goal is to build something that is just good enough to be useful. It doesn't need to be even a full fledged conversational LLM. It could just be an LM. Lots of tasks don't require knowledge to be encoded into the model. For example "summarize my emails" is a task that can be done with a fairly small model trained on just basic text. There are also unexplored avenues. For example, if I had the hardware to do this, I would basically take an early version of GPT, and then start training it on additional data, and when the training run completes, I would diff the model with the original version, and use that as the training set of another model. Basically build a model on top of GPT that can automatically adjust parameter weights and encode it into the model, thus giving it persistent memory.
    - BeFlatXIII 2y ago
      LLMs themselves won't go away, but they may not last as a sustainable business.
- player1234 2y ago
  Luddite, didn't you read in this very thread someone's dad is using it for excel macros! That is at least worth 500 trillions!
pjc50 2y ago
If its proponents are taken at face value, AGI is a threat.

AGI is a technology or a feature, not a product. ChatGPT is a product. They need some more products to pay for one of the most expensive technologies ever (to not be delivered yet).

leptons 2y ago

It's a shame that $1 trillion is being poured into AI so quickly, but fusion research has only seen a fraction of that over many decades.

andsoitis 2y ago
Isn’t that a reflection, to some (major?) extent, of the general sense of likelihood of breakthrough?
- xmprt 2y ago
  I used to think like this but after seeing the amount of money invested into crypto companies which most average people could have quickly dismissed as irrelevant, I'm not sure VCs are a good judge of value.
  - 9rx 2y ago
    That's not an entirely fair take. Once upon a time it was believed that cryptocurrencies would be able to supplant payment processors (i.e. VISA, Master Card) We even had a rollout of crypto ATMs to utilize the belief in that network. That is a money printing business, so it wasn't unreasonable for VC money to chase it. Sure, those intimately familiar with the technology may have always known that it wouldn't work, but the average person absolutely did not. They only came to learn that it wouldn't work as the cryptocurrency ventures proved that it wouldn't work. The pivot to being some kind of tool to store wealth, or a way to trade cartoon monkeys, or whatever as a last ditch effort to save the companies that failed to turn it into payment processing technology was more obviously irrelevant, but that came much, much later. The VCs were desperately seeking an exit by that point.
- j_maffe 2y ago
  No it's the perception of likelihood of breakthrough
- caseyy 2y ago
  It’s why we have warnings to investors saying past performance is not an indicator of future returns.
- plorg 2y ago
  Maybe it means investors think it will happen, maybe it means investors simply think it will be profitable. It could even be investors seeking a market that has juice long after ZIRP dried up.
- rchaud 2y ago
  Capital gains on AI investment doesn't need a breakthrough. I'm sure pharma companies and others are building their own custom models to assist R&D, but those aren't the ones sinking the truly huge sums into consumer-facing LLMs.
- leptons 2y ago
  We already know fusion exists and happens and has been happening for billions of years - it's well understood. It may not be simple to reproduce and sustain on earth, but there is a fairly clear path to get there. There has never been an AGI, and there's no clear path to get there. And no, LLMs are not bringing us any closer to a real AGI, no matter how much Sam Altman wants to try to redefine what "AGI" means so that he can claim he has it. The difference between AGI funding and fusion funding is the people hawking AGI are better liars and the people funding are far more gullible than the people trying to make fusion power a reality.
  - Nevermark 2y ago
    We already know fusion has been a grind. For decades. We already know that computation has grown exponentially in scale and cost, for decades. With breakthrough software tech showing up reliably as it harnesses greater computational power, even if unpredictably and sparsely. If we view AI across decades, it has reliably improved. Even through its “winters”. It just took time to hit thresholds of usefulness making relatively smooth progress look very random from a consumers point of view. As computational power grows, and current models become more efficient, the hardware overhang for the next major step is growing rapidly. That just hasn’t been the case for fusion.
    - leptons 2y ago
      >We already know fusion has been a grind. For decades. Let's also not forget that "AI" has been grinding on for at least half a century, with nothing even coming close to AGI. No, LLMs are not AGI and cannot be AGI. They may be a component of an eventual AGI in the future, but we're still far, far away from artificial general intelligence. LLMs are not intelligent, but they do fool a lot of people. With the paltry amount of money spent on fusion, it's no wonder it's been delayed for so long. Maybe try putting the $1 Trillion into it in the short time "AI" has had its windfall, and let's see what happens with fusion. >If we view AI across decades, it has reliably improved That's your opinion. It's still not much better than Eliza. And LLMs lie half the time, just to make the user happy hearing what they want to hear. >As computational power grows, and current models become more efficient, the hardware overhang for the next major step is growing rapidly. And it's still going to "hallucinate" and give wrong answers a lot of the time because the fundamental tech behind LLMs is a bit flawed and is in no way even close to AGI. >That just hasn’t been the case for fusion. And yet we have actual fusion reactions happening now, sustained for 22 minutes. We're practically on the threshold of having fusion power, but still there is nowhere near the money being spent on it compared to LLMs, which won't be capable of AGI in spite of the marketing line you've been told.
imafish 2y ago
Unfortunately VC investments rely on FOMO, and there is no current huge breakthrough in fusion research that they are afraid of missing out on.
Tenoke 2y ago
Much, much less was being poured into AI until it started to have some returns.
- leptons 2y ago
  What returns? OpenAI is still not profitable.
csharpminor 2y ago
If you believe the AGI thesis, this is a stepping stone to unlocking fusion and other scientific advances.
immibis 2y ago
Once we unlock fusion, humanity starts down a millenia-long path to making the planet permanently uninhabitable (not just temporarily like with climate change) by turning all the water into bitcoins.
- orbifold 2y ago
  I don’t think that is true, there is only so much heavy water on the planet and doing hydrogen fusion is not all that economical.
- leptons 2y ago
  "There are 1,335,000,000,000,000,000,000 liters of water in all the oceans." "A large fusion power plant, generating 1,500 megawatts of electricity, would consume approximately 600 grams of tritium and 400 grams of deuterium each day. A 1,000 MW coal-fired power plant requires 2.7 million tonnes of coal per year, while a similar fusion plant would need only 250 kg of fuel per year (half deuterium, half tritium)." You do the math. I'm personally not worried about running out of water due to fusion.
bookofjoe 2y ago
Gemini 10.0 Wait for it
BeFlatXIII 2y ago
It'll be fun to point and laugh at everyone when the bubble bursts.
- ActorNightly 2y ago
  The bubble won't burst. LLMs are very useful. There just hasn't been enough optimization for general use case, which is kinda how all tech starts. Eventually someone will make a more optimized lower parameter model that has decent accuracy, and can be run on lower grade hardware, that will be a pretty good assistant with manual wrappers around automation.
  - leptons 2y ago
    This is the same llm-splaining that practically every enthusiast does. It's either "more hardware" or "more optimization" will solve all the problems and bring us into a new era of blah blah blah. But it won't. It won't solve LLM slop or hallucinations, which are inherent to how LLMs work, and why that bubble is already looking pretty unstable.
    - ActorNightly 2y ago
      The goal isn't to build perfect assistants with LLM. The goal is to build something that is just good enough to be useful. It doesn't need to be even a full fledged conversational LLM. It could just be an LM. Lots of tasks don't require knowledge to be encoded into the model. For example "summarize my emails" is a task that can be done with a fairly small model trained on just basic text. There are also unexplored avenues. For example, if I had the hardware to do this, I would basically take an early version of GPT, and then start training it on additional data, and when the training run completes, I would diff the model with the original version, and use that as the training set of another model. Basically build a model on top of GPT that can automatically adjust parameter weights and encode it into the model, thus giving it persistent memory.
  - BeFlatXIII 2y ago
    LLMs themselves won't go away, but they may not last as a sustainable business.
player1234 2y ago
Luddite, didn't you read in this very thread someone's dad is using it for excel macros! That is at least worth 500 trillions!

pjc50 2y ago

If its proponents are taken at face value, AGI is a threat.