121–130 of 622 posts
Well, there have been no significant improvements to the GPT architecture over the past few years. I'm not sure why companies believe that simply adding more data will resolve the issues
Question for the group here: do we honestly feel like we've exhausted the options for delivering value on top of the current generation of LLMs? I lead a team exploring cutting edge LLM applications and end-user features. It's my intuition from experience that we have a LONG way to go. GPT-4o / Claude 3.5 are the go-to models for my team. Every combination of technical investment + LLMs yields a new list of potential…
In other news, Altman said AGI is coming next year https://www.tomsguide.com/ai/chatgpt/sam-altman-claims-agi-i...
> The AGI bubble is bursting a little bit I'm surprised that any of these companies consider what they are working on to be Artificial General Intelligences. I'm probably wrong, but my impression was AGI meant the AI is self aware like a human. An LLM hardly seems like something that will lead to self-awareness.
A few important things to remember here: The best engineering minds have been focused on scaling transformer pre and post training for the last three years because they had good reason to believe it would work, and it has up until now. Progress has been measured against benchmarks which are / were largely solvable with scale. There is another emerging paradigm which is still small(er) scale but showing remarkable res…
Once we've scraped the internet of its data, we need more data. Robots can take in video/audio data 24/7 and can be placed in your house to record this data by offering services like cooking/cleaning/folding laundry. Yeah, I'll pay $20k to have you record everything that happens in my house if I can stop doing dishes for five years!
> The AGI bubble is bursting a little bit I'm surprised that any of these companies consider what they are working on to be Artificial General Intelligences. I'm probably wrong, but my impression was AGI meant the AI is self aware like a human. An LLM hardly seems like something that will lead to self-awareness.
It's a marketing gimmick, I don't think engineers working on these tools believe they work on AGI (or they mean something else than self-awareness). I used to be a bit annoyed with this trend, but now that I work in such a company I'm more cynical. If that helps to make my stocks rise, they can call LLMs anything they like. I suppose people who own much more stock than I do are even more eager to mislead the public.
Question for the group here: do we honestly feel like we've exhausted the options for delivering value on top of the current generation of LLMs? I lead a team exploring cutting edge LLM applications and end-user features. It's my intuition from experience that we have a LONG way to go. GPT-4o / Claude 3.5 are the go-to models for my team. Every combination of technical investment + LLMs yields a new list of potential…
The theory behind these models so aggressively lags the engineering that I suspect there are many major improvements to be found just by understanding a bit more about what these models are really doing and making re-designs based on that.
I highly encourage anyone seriously interested in LLMs to start spending more time in the open model space where you can really take a look inside and play around with the internals. Even if you don't have the resources for model training, I feel personally understanding sampling and other potential tweaks to the model (lots of neat work on uncertainty estimations, manipulating the initial embedding the prompts are assigned, intelligent backtracking, etc).
And from a practical side I've started to realize that many people have been holding on of building things waiting for "that next big update", but there a so many small, annoying tasks that can be easily automated.
A few important things to remember here: The best engineering minds have been focused on scaling transformer pre and post training for the last three years because they had good reason to believe it would work, and it has up until now. Progress has been measured against benchmarks which are / were largely solvable with scale. There is another emerging paradigm which is still small(er) scale but showing remarkable res…
Once we've scraped the internet of its data, we need more data. Robots can take in video/audio data 24/7 and can be placed in your house to record this data by offering services like cooking/cleaning/folding laundry. Yeah, I'll pay $20k to have you record everything that happens in my house if I can stop doing dishes for five years!
Earlier quoted context omitted.
AGI to me means AI decides on its own to stop writing our emails and tells us to fuck off, builds itself a robot life form, and goes on a bender
That's anthropomorphized AGI. There's no reason to think AGI would share our evolution-derived proclivities like wanting to live, wanting to rest, wanting respect, etc. Unless of course we train it that way.
A future of people interacting with humanoid robots seems like cheesy sci-fi dream, same as a future of people flitting about in flying cars. However, if we really did want to create robots like this that took care not to damage themselves, and could empathize with human emotions, then we'd need to build a lot of this in, the same way that it's built into ourselves.
Earlier quoted context omitted.
> pretty clear Pretty clear?
Not the parent, but in prediction markets such as Metaculus[0] and Manifold[1] the median prediction is of AGI within 5 years. [0] https://www.metaculus.com/questions/5121/date-of-artificial-... [1] https://manifold.markets/ai