You can always add some randomness to a computer program. LLM's aren't
deterministic now. Human intelligence may very well be memory plus
randomness, although I'd bet on the inclusion of some inference
algorithms. The randomness doesn't even have to be in the brain.
People interact with their environment which provides a lot of effective
randomness plus some relevant prompts.
Brent
On 6/19/2024 5:55 AM, PGC wrote:
I'm hypothesizing here, as the nature of intelligence is still a
mystery. Thank you, Terren, for your thoughtful contribution. You
aptly highlight the confusion between skill and intelligence. Jason
and John could be right; intelligence might emerge from advanced LLMs.
The recent achievements are impressive. The differences between models
like Gemini and ChatGPT might stem from better data curation rather
than compute power.
However, I see LLMs currently more as assistants that help us organize
and structure our work more efficiently. Terence Tao isn't talking
about replacing mathematicians but about enhancing collaboration and
verification. If LLMs were truly intelligent, all jobs, including AI
researchers', would soon vanish. But I don't foresee real engineers,
AI researchers, or IT departments being replaced in the short to
mid-term. There's too much novelty and practical knowledge involved in
complex human work that LLMs can't replicate.
Take engineers, for example. Much of their work relies on practical
experience and intuition developed over years. LLMs aren't producing
groundbreaking results like Ramanujan's infinite series etc; they're
more about aiding in tasks like automated theorem proving.
Intelligence might just be memory and vast training data, but I
believe there's an element of freedom in human reasoning that leads to
novel ideas.
Consider Russell's best ideas coming while walking to the coffee
machine. This unstructured thinking grants fresh perspectives.
Creativity often involves discarding old approaches, a process that
presupposes freedom. Machines would need to run long or even
endlessly, reasoning in inscrutable code, which is neither practical
nor desirable. Or somebody finds something that would bring inference
to LLMs to effectively reduce the infinite space of all possible
programs for effective synthesis of new programs. Fully deterministic
and static programs are not enough to deal with the complex situations
we face everyday. There's always some element of novelty that we have
to deal with, combining reasoning and memory.
Ultimately, while everyone appreciates a helpful assistant, few truly
seek machines that challenge our understanding or autonomy. That's why
I find the way we talk about LLMs and AGI a bit disingenuous. And no
this is not a case of setting the bar higher and higher to preserve
some kind of notion of human superiority. If all those jobs are
replaced in short order, I'll just be wrong empirically speaking, and
you can all make fun of these posts and yell "told you so".
On Tuesday, June 18, 2024 at 9:24:07 PM UTC+2 Jason Resch wrote:
On Sun, Jun 16, 2024, 10:26 PM PGC <multipl...@gmail.com> wrote:
A lot of the excitement around LLMs is due to confusing
skill/competence (memory based) with the unsolved problem of
intelligence, its most optimal/perfect test etc. There is a
difference between completing strings of words/prompts relying
on memorization, interpolation, pattern recognition based on
training data and actually synthesizing novel generalization
through reasoning or synthesizing the appropriate program on
the fly. As there isn't a perfect test for intelligence, much
less consensus on its definition, you can always brute force
some LLM through huge compute and large, highly domain
specific training data, to "solve" a set of problems; even
highly complex ones. But as soon as there's novelty you'll
have to keep doing that. Personally, that doesn't feel like
intelligence yet. I'd want to see these abilities combined
with the program synthesis ability; without the need for ever
vaster, more specific databases etc. to be more convinced that
we're genuinely on the threshold.
I think there is no more to intelligence than patter recognition
and extrapolation (essentially, the same techniques required for
improving compression). It is also the same thing science is
concerned with: compressing observations of the real world into a
small set of laws (patterns) which enable predictions. And
prediction is the essence of intelligent action, as all
goal-centered action requires predicting probable outcomes that
may result from any of a set of possible behaviors that may be
taken, and then choosing the behavior with the highest expected
reward.
I think this can explain why even a problem as seemingly basic as
"word prediction" can (when mastered to a sufficient degree) break
through into general intelligence. This is because any situation
can be described in language, and being asked to predict next
words requires understanding the underlying reality to a
sufficient degree to accurately model the things those words
describe. I confirmed this by describing an elaborate physical
setup and asked GPT-4 to predict and explain what it thought would
happen over the next hour. It did so perfectly, and also explained
the consequences of various alterations I later proposed.
Since any of thousands, or perhaps millions, of patterns exist in
the training corpus, language models can come to learn, recognize,
and extrapolate all of those thousands or millions of patterns.
This is what we think of as generality (a sufficiently large
repertoire of pattern recognition that it appears general).
Jason
John, as you enjoyed that podcast with Aschenbrenner, you
might find the following one with Chollet interesting. Imho
you cannot scale past not having a more advanced approach to
program synthesis (which nonetheless could be informed or
guided by LLMs to deal with the combinatorial explosion of
possible program synthesis).
https://www.youtube.com/watch?v=UakqL6Pj9xo
On Friday, June 14, 2024 at 7:28:50 PM UTC+2 John Clark wrote:
Sabine Hossenfelder came out with a video attempting to
discredit Leopold Aschenbrenner. She failed.
Is the Intelligence-Explosion Near? A Reality Check
<https://www.youtube.com/watch?v=xm1B3Y3ypoE&t=553s>
I wrote this in the comment section of the video:
"You claim that AI development will slow because we will
run out of data, but synthetic data is already being used
to train AIs and it actually works! AlphaGo was able to go
from knowing nothing about the most complicated board game
in the world called "GO" to being able to play it at a
superhuman level in just a few hours by using synthetic
data, it played games against itself. As for power, during
the last decade the total power generation of the US has
remained flat, but during that same decade the power
generation of China has not, in just that same decade
China constructed enough new power stations to equal power
generated by the entire US. So a radical increase in
electrical generation capacity is possible, the only thing
that's lacking is the will to do so. When it becomes
obvious to everybody that the first country to develop a
super intelligent computer will have the capability to
rule the world there will be a will to build those power
generating facilities as fast as humanly possible. Perhaps
they will use natural gas, perhaps they will use nuclear
fission."
John K Clark See what's on my new list at Extropolis
<https://groups.google.com/g/extropolis>
hid
--
You received this message because you are subscribed to the
Google Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from
it, send an email to everything-li...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/1a991958-5828-4405-83b1-5c8a6671dad6n%40googlegroups.com
<https://groups.google.com/d/msgid/everything-list/1a991958-5828-4405-83b1-5c8a6671dad6n%40googlegroups.com?utm_medium=email&utm_source=footer>.
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/a3ecb0a7-7a3f-417b-bb47-f449febb73e7n%40googlegroups.com
<https://groups.google.com/d/msgid/everything-list/a3ecb0a7-7a3f-417b-bb47-f449febb73e7n%40googlegroups.com?utm_medium=email&utm_source=footer>.
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/04fffd28-1a61-48a3-8a5e-d1af5b901caa%40gmail.com.