131–140 of 326 posts
>“That language models can do these sort of things was never discussed in any literature that I’m aware of," I had previously the expectation that unpredictable emergent behavior would exist in any sufficiently complex system? Based on layman's readings in chaos and complexity theory.
Earlier quoted context omitted.
What they said 10 years ago was correct. It did hit society like a sledge hammer. Machine learning basically took over the AI space and penetrated the consumer space with applications that were all but impossible in the previous decade. There's AI chips in smart phones now. What you're seeing here with LLMs is sledge hammer number 2. It's understandable how most people don't notice the sledge hammer. The decade prior…
Sledge hammer #1 (voice assistants, AI chips in phones) didn’t cause unemployment. It was at the level of new features and capabilities. Sledge hammer #2 is aimed squarely at “white collar” work without much in the way of bounds to its capabilities.
Earlier quoted context omitted.
I’m looking forward to the “alternative history” SF novels of the future where AGI never materialized.
Dune is a great example of that! After their AGI goes sideways (in the Butlerian Jihad) human civilization shifts entirely to using human supercomputers (Mentats and the Bene Gesserit) juiced up on mental enhancement drugs (Spice).
Earlier quoted context omitted.
The same was said 10 years ago. It's astonishing what can be done, but you can already see fundamental limits. I think it will raise productivity for some tasks, but not fundamentally change society.
The A.I. Monte Carlo Ray Tracing Hallucination Engine can change society by showing as-is and to-be next state. Two use-cases: new infrastructure installation or upgrade and time interval inspection tracking ideal-case vs real world condition. Suppose a lazy contractor skips steps and cuts corners, or a pathological contractor builds the thing and pulls the thing apart over and over again when all that was needed was…
Earlier quoted context omitted.
there is no channel for uncertainty. LLMs of this type will just start making up shit when they dont know something. because they simply generate the most probable next token based on previous x tokens. this is not fixable. this alone makes these LLMs practically unusable in vast majority of real-world applications where you would otherwise imagine this tech to be used.
yea its a simulator of human text on the internet for instance, your comment confidently states this is unfixable - presumably based on the frequency you've seen similar text on the internet. why should anyone believe the veracity of your statement? These things didn't have any of these emergent capabilities one year ago, why are you so sure you understand their nature one year from now?
no its because GPT is based on transformers.
Earlier quoted context omitted.
there is no channel for uncertainty. LLMs of this type will just start making up shit when they dont know something. because they simply generate the most probable next token based on previous x tokens. this is not fixable. this alone makes these LLMs practically unusable in vast majority of real-world applications where you would otherwise imagine this tech to be used.
It's not clear why this would be a fundamental limit rather than a design flaw that will eventually be solved.
Earlier quoted context omitted.
there is no channel for uncertainty. LLMs of this type will just start making up shit when they dont know something. because they simply generate the most probable next token based on previous x tokens. this is not fixable. this alone makes these LLMs practically unusable in vast majority of real-world applications where you would otherwise imagine this tech to be used.
this is absolutely not a fundamental limit but simply a hard challenge. Approaches exist and it is an active field of research where we do make progress.
Earlier quoted context omitted.
there is no channel for uncertainty. LLMs of this type will just start making up shit when they dont know something. because they simply generate the most probable next token based on previous x tokens. this is not fixable. this alone makes these LLMs practically unusable in vast majority of real-world applications where you would otherwise imagine this tech to be used.
What if LLM knowledge will expand over time to be sufficient for certain real-world application?
Earlier quoted context omitted.
When I was a bored and under-challenged student in the early days of university, one of my tests during a programming exam was to write a program that spit out the factorial of an input integer. For shits and giggles, I just just wrote a series of `if...then` statements for 0 to 5, and only actually calculated the factorial for inputs >=6. I passed after the examiner just input 3 and 5 and was satisfied that the outp…
Doing a table lookup for common cases and computing for the less common ones is perfectly valid!
Earlier quoted context omitted.
Sledge hammer #1 (voice assistants, AI chips in phones) didn’t cause unemployment. It was at the level of new features and capabilities. Sledge hammer #2 is aimed squarely at “white collar” work without much in the way of bounds to its capabilities.
We can't trust it. That's a pretty hard bound.
Trust is only a bound for certain areas, and this bound is eroding.