Yet another Hacker News

The Price of Intelligence Is Falling Rapidly

16 points gotmedium 1 hour ago 13 comments

silverFork 56 minutes ago | parent

Price of human intelligence is also falling rapidly. I suspect looking pretty is going to go up in value more rapidly.

0c3ca83 21 minutes ago | parent

Talking pretty, too -- which takes some intelligence, but not too much. The era of the awkward nerd is over. The world is returning to its natural order, with the slimy salesman in the lead.

sph 7 minutes ago | parent

Price of human intelligence is going up actually, because there's ever-decreasing amount of it, and we've invented a cheap substitute that makes people even stupider.

hyperionultra 27 minutes ago | parent

Well, since llms need matterial for self-education, someone still had to be inteligent enough. Or llms will freeze their “intelect” at some point.

hlynurd 20 minutes ago | parent

Well, someone. But not necessarily anyone else.

BenzeneDream 17 minutes ago | parent

Why would it freeze if they can create more training material. Its already been shown synthetic training data is just as good if not better.

timcobb 8 minutes ago | parent

> Its already been shown synthetic training data is just as good if not better.

Source, please! I'm a machine learning layperson but how can this not be overfitting?

azakai 6 minutes ago | parent

For example, an LLM can prove a new theorem in math, verify it in Lean (so it is definitely true), and then prove more things based on that.

AI-generated data might not always be that useful, but at least in this case it obviously is.

hiddencost 9 minutes ago | parent

Surprisingly, this isn't necessarily true.

To take an intuition: there are lots of problems that are easier to pose and to check than they are to solve, in mathematics.

bonoboTP 8 minutes ago | parent

No, they can just solve tasks verifiably. Reinforcement learning from verifiable rewards. Pure next-token supervision is only in the pre-training phase for modern LLM agents. They can also define new RL environments and pose new challenges to themselves and train themselves to solve them faster. Yes, at some point some human steering and judgment comes in as to what sorts of tasks to train towards. But there is no obvious taper-off at human level.

2OEH8eoCRo0 18 minutes ago | parent

A race to the bottom on price is very good for profits!

129348681 6 minutes ago | parent

The actual report is by Epoch AI, which is financed by totally independent funds like (Eric) Schmidt Sciences.

Professor Tabarrok uncritically praises the report, which might lead to the conclusion that either human intelligence or moral standards are "falling rapidly".

serious_angel 2 minutes ago | parent

These ML/LLM/AI are systems where no one will ever be accountable for and guarantee its stability, safety, and its value.

ML/LLM/AI may be useful, but it must always be considered unauthorized, mistrust, experimental, unverified, always probable to fail on each iteration. Its data must always be accountably supervized by an alive human, who realizes responsibility, life, pain, art, effort, love... Always, even at 2999+ years/times. No life/pain/art/love is no sense/trust/confidence/respect/safety.

That is, it may be fun to play with as a toy, as any ML/LLM/AI based system, where no trust exist but only probability, entropy, chance, luck... but nothing serious, time-rewarding, or humane enough to appreciate and love it in the long run.

An actual Developer is someone who prioritizes thinking and solution finding relying on something deterministic and accountable.

All systems based on LLM, ML, or anything that is algorithmically based on feedback system, learning, training, is indeterministic, and hence, always inaccurate, unreliable, and irresponsible systems, by design.

The process of art-working, coding, programming, development, solution finding... that feel of an achievement. This is what I understand, too. There's no achievement in anything AI-driven - it based on luck/moment/probability/chance.

"Oh My God! Please, all Gods, make this AI output the finally correct code! Please!"

Fudge it.

And you also pay for it. You pay and depend on it, on the the luck, for the chance, for the probability someone controls for you over an API or algorithm you don't even know or cannot verify due its "self-training" design, or it being close-sourced.

> ...you're gatekeeping programming... you are not seeing the big picture... you are missing opportunities...

No, sorry, never. I do not agree with that. These "responses" are coming from infants, immature, children, I believe, sorry... from those who could not care less about the development, responsibility, the overall process of the life of us, people, to find solutions, to reach the goals, to treasure achievements, to write our history with actual names of people on papers sign on papers by hand... for what will leave after us, with art, statues, quotes... in the infinite history of the ineffably magnificent world...

Your name should be in the code you write alone or with a team, your commits, solutions. Your work must be acknowledged and kept in the history. You should be always a respectable alive artist, who is responsible for their work. Our history must evolve, be preserved, respected, appreciated, and shine with glory of our achievements with human names in it for the incredibly marvelous future of opportunities, discoveries, love, and purely magnificent marvel.

There's whole space out there, still, and we must add and keep human names in the history cherished.

> Machines should work. People should think. ~ IBM