4 ms·

Gpt-4 hallucinates meaningfully less than gpt-3. There is more evidence in favor of “more improvements to come” than “ai winter approaches”. A lot smart peopl

Gpt-4 hallucinates meaningfully less than gpt-3. There is more evidence in favor of “more improvements to come” than “ai winter approaches”. A lot smart people seem to be saying that the existing approaches have room to improve simply by training with more data. Based on what I’ve read and roughly speculating, it looks there is easily enough existing data for gpt-5 and probably a few more versions after. I’m not sure how easy it will be to legally acquire all of the available data though.

antupis 4y ago
I think good annotation is much harder than obtaining data. But now Reddit, Twitter and Quora will realise what kind goldmine their data is they might close easy access to it.
- bpodgursky 4y ago
  None of the GPT models rely on annotated or classified data. It's unsupervised.
  - IanCal 4y ago
    The final training steps of gpt 3.5 did, and we assume the same for 4. The RLHF step.
    - whimsicalism 4y ago
      I would say semi-supervised, RLHF is not quite supervised learning. None of this data is from Reddit, Twitter, etc.
- wordpad25 4y ago
  It’s too easy to scrape them, but even that is already done by many intermediary companies all too happy to resell that data.
  - antupis 4y ago
    Yup but you still need up to date data and now companies are realising what kinda gold mine random forum posts about nightlife in Glasgow are they will start monetising it and probably try at least kill middle market.
p-e-w 4y ago
> There is more evidence in favor of “more improvements to come” than “ai winter approaches”. Not to mention that we haven't really seen what GPT-4 can do to the world yet. Most people are still treating it as a cool toy, when it can probably replace 20% of standard office workers across all professions with the capabilities it has today. I mean, ChatGPT is out-diagnosing medical professionals and out-advising experienced lawyers in some cases. How do you think it would fare against the average (essentially untrained) customer support representative? The major effects are yet to come, and there doesn't even have to be a GPT-5 for that to happen – the world just needs to fully realize what GPT-4 already is.
dragonwriter 4y ago
> There is more evidence in favor of “more improvements to come” than “ai winter approaches”. Every “AI Winter” has seen continuing improvements, its just that the pace and impact of the improvements fell short of the accelerating anticipation driven by the preceding AI hype cycle and the demands of those investing capital into AI. If there is another AI winter, it will be the same. “More improvements to come” is not the opposite of “AI winter approaches”; its not as if expert systems stopped advancing after the second AI winter (or neural nets after the first.)
- Nevermark 4y ago
  > “More improvements to come” is not the opposite of “AI winter approaches”; its not as if expert systems stopped advancing after the second AI winter (or neural nets after the first.) True. And its also true that even in that nuanced form, a slowdown not a stop, there is no evidence of an imminent AI winter. Major progress in model behavior is happening on the scale of months. Happening often based on relatively obvious or routine improvements, such as more data, more kinds of data, more computation, expanding the size of prompt windows, tuning of model architectures, etc. And not only businesses, but consumers, are getting access to these tools in ways that are increasing demand for more improvements. And for these tools to be more available to all kinds of areas of work, creativity, and leisure. So the resources that are being brought to bear on routine and more fundamental improvements are growing at an exceptional pace. No sign of a saturation of progress in sight. Quite the reverse.
f6v 4y ago
> A lot smart people seem to be saying that the existing approaches have room to improve simply by training with more data. There’re also many “smart people” saying that simply having more data doesn’t turn a machine that can’t reason into one that can.
- pmoriarty 4y ago
  "simply having more data doesn’t turn a machine that can’t reason into one that can" But that's exactly what's been happening. LLM's are a triumph of "big data" used by dumb algorithms to create a reasoning machine.
  - trompetenaccoun 4y ago
    A reasoning machine that can't reason, makes up fake evidence and doesn't follow simple logic.
    - SanderNL 4y ago
      Are we talking about humans or..
  - f6v 4y ago
    > dumb algorithms I’m not an expert and haven’t seen architecture of ChatGPT. But it’s probably far from dumb. You can’t simply put more data into any algorithm and expect improvement. You need a model to be able to take advantage of the data and train in reasonable time. That’s exactly what’s been happening in ML field.
    - dwaltrip 4y ago
      A huge portion of the improvements in each successive version of gpt came from increasing the model size and the amount of training data. The research trends indicate that there still are additional improvements to be found by continuing to scale things up, especially when it comes to the amount of training data. Not indefinitely, obviously. But we haven’t exhausted the available data sources yet, and the curves for prediction error vs. amount of training data strongly imply using this additional data would result in increased performance. Do you have specific reasons for thinking that gpt-4 is near the end of this trend? I haven’t seen any compelling arguments for this.

antupis 4y ago

I think good annotation is much harder than obtaining data. But now Reddit, Twitter and Quora will realise what kind goldmine their data is they might close easy access to it.

bpodgursky 4y ago
None of the GPT models rely on annotated or classified data. It's unsupervised.
- IanCal 4y ago
  The final training steps of gpt 3.5 did, and we assume the same for 4. The RLHF step.
  - whimsicalism 4y ago
    I would say semi-supervised, RLHF is not quite supervised learning. None of this data is from Reddit, Twitter, etc.
wordpad25 4y ago
It’s too easy to scrape them, but even that is already done by many intermediary companies all too happy to resell that data.
- antupis 4y ago
  Yup but you still need up to date data and now companies are realising what kinda gold mine random forum posts about nightlife in Glasgow are they will start monetising it and probably try at least kill middle market.

p-e-w 4y ago

> There is more evidence in favor of “more improvements to come” than “ai winter approaches”. Not to mention that we haven't really seen what GPT-4 can do to the world yet. Most people are still treating it as a cool toy, when it can probably replace 20% of standard office workers across all professions with the capabilities it has today. I mean, ChatGPT is out-diagnosing medical professionals and out-advising experienced lawyers in some cases. How do you think it would fare against the average (essentially untrained) customer support representative? The major effects are yet to come, and there doesn't even have to be a GPT-5 for that to happen – the world just needs to fully realize what GPT-4 already is.

dragonwriter 4y ago

> There is more evidence in favor of “more improvements to come” than “ai winter approaches”. Every “AI Winter” has seen continuing improvements, its just that the pace and impact of the improvements fell short of the accelerating anticipation driven by the preceding AI hype cycle and the demands of those investing capital into AI. If there is another AI winter, it will be the same. “More improvements to come” is not the opposite of “AI winter approaches”; its not as if expert systems stopped advancing after the second AI winter (or neural nets after the first.)

Nevermark 4y ago
> “More improvements to come” is not the opposite of “AI winter approaches”; its not as if expert systems stopped advancing after the second AI winter (or neural nets after the first.) True. And its also true that even in that nuanced form, a slowdown not a stop, there is no evidence of an imminent AI winter. Major progress in model behavior is happening on the scale of months. Happening often based on relatively obvious or routine improvements, such as more data, more kinds of data, more computation, expanding the size of prompt windows, tuning of model architectures, etc. And not only businesses, but consumers, are getting access to these tools in ways that are increasing demand for more improvements. And for these tools to be more available to all kinds of areas of work, creativity, and leisure. So the resources that are being brought to bear on routine and more fundamental improvements are growing at an exceptional pace. No sign of a saturation of progress in sight. Quite the reverse.

f6v 4y ago

> A lot smart people seem to be saying that the existing approaches have room to improve simply by training with more data. There’re also many “smart people” saying that simply having more data doesn’t turn a machine that can’t reason into one that can.

pmoriarty 4y ago
"simply having more data doesn’t turn a machine that can’t reason into one that can" But that's exactly what's been happening. LLM's are a triumph of "big data" used by dumb algorithms to create a reasoning machine.
- trompetenaccoun 4y ago
  A reasoning machine that can't reason, makes up fake evidence and doesn't follow simple logic.
  - SanderNL 4y ago
    Are we talking about humans or..
- f6v 4y ago
  > dumb algorithms I’m not an expert and haven’t seen architecture of ChatGPT. But it’s probably far from dumb. You can’t simply put more data into any algorithm and expect improvement. You need a model to be able to take advantage of the data and train in reasonable time. That’s exactly what’s been happening in ML field.
  - dwaltrip 4y ago
    A huge portion of the improvements in each successive version of gpt came from increasing the model size and the amount of training data. The research trends indicate that there still are additional improvements to be found by continuing to scale things up, especially when it comes to the amount of training data. Not indefinitely, obviously. But we haven’t exhausted the available data sources yet, and the curves for prediction error vs. amount of training data strongly imply using this additional data would result in increased performance. Do you have specific reasons for thinking that gpt-4 is near the end of this trend? I haven’t seen any compelling arguments for this.