I think its down to language and domain more than tools.
No model ive tried can write, usefully debug or even explain cmake. (It invents new syntax if it gets stuck, i often have to prompt multiple AI to know if even the first response in the context was made-up)
My luck with embedded c has been atrocious for existing codebase (burning millions of tolkens), but passable for small scripts. (Arduino projects)
My experience with python is much better. Suggesting relevant libraries and functions, debugging odd errors, or even making small script on its own. Even the original github copilot which i got access to early was excellent on python.
Alot of people that seem to have fully embraced agentic vibe-coding seem to be in the web or node.js domain. Which I've not done myself since pre-AI.
I've tried most (free or trial) major models or schemes in hope that i find any of them useful, but not found much use yet.
Slow and ram heavy. Zed feels refreshingly snappy compared to vscode even before adding plugins. And why does desktop application need to use interpreted programming languages?
zed was just a fast and simple replacement for Atom (R.I.P) or vscode. Then they put AI on top when that showed up. I don't care for it, and appreciate a project like this to return the program to its core.
The entire argument is that thinking it's intelligent or a person makes us missuse the tool in dangerous ways. Not to make us feel better; but to not do stupid things with them.
As a tool its useful yes, that is not the issue;
- theyre used as phycologist and life coaches.
- judges of policy and law documents
- writers of life affecting computer systems.
- Judges of job applications.
- Sources of medical advice,
- legal advisors
- And increasingly as a thing to blame when any of above goes awry.
If we think of llms as very good text writing tools, the responsibility to make "intelligent" decisions and more crucially take responsibility for those decisions remains on real people rather than dice.
But if we think of them as intelligent humans, we making a fatal misjudgement.
> I think winning a Go or a chess competition does demonstrate intelligence.
Chess is a simple alfa beta pruned minmax seaech tree. If that's intelligent then a drone flight controller or a calculator is aswell.
> association and recall, not for performing complex logical transformations.
By that definition humans doing chess aren't as intelligent as a computer doing chess, since high level chess is heavily reliant on memory and recall of moves and progressions.
I did not share any definitions, only vague opinions. Not that I'd know what it means for a definition to "fall apart".
And the specific bit you cite is barely even a vague opinion; it is my interpretation of the show "Jeopardy!" based on the Wiki article (I've never seen a single episode, wasn't really a thing where I'm from):
> Specifically because it reads like it's about (...) knowledge: it tests for association and recall (...)
Also:
> By that definition humans doing chess aren't as intelligent as a computer doing chess, since high level chess is heavily reliant on memory and recall of moves and progressions.
Yes, I did find this really quite disappointing and disillusioning when I first learned about it. A colleague of mine even straight up quit competitive chess over it.
> it is my interpretation of the show "Jeopardy!" based on the Wiki article
You are spot on though. I mostly wanted to argue that no decent distinction can be made here.
> I did find this really quite disappointing and disillusioning when I first learned about it
ye... same here.
---
I'm personally in the camp that "intelligence" is a human concept. A metric to compare humans. Applying it to computers makes us anthropomorphism computers and think of them as people. Thinking of LLMs as people makes us trust them with bad things.
So we should call them impressive, fluent, fast, useful, good at tasks. Computers already beat us at most math, statistics, searching for information, spacial visualization, information recollection, lossless communication. LLMs just adds to that list, but does nothing new to make the word "intelligent" applicable. Even if we reach the AGI singularity; thinking of them as humans or using human terminology to describe them is a fatal error.
(Destroying earth to make paperclips is arguably the least intelligent thing you could ever do.)
> there’s no obvious algorithm for discovering mathematical proofs or solving difficult problems that could be implemented in a classical, pre-Gen AI computer program.
Fundamentally opposite. Computer algorithms have been part of math research since they where invented, and mathematical proof algorithms are widespread and excellent.
The llms that are now "intelligent enough to do maths" are just trained to rephrase questions into prolog code.
For the OpenAI case, its unclear. They've not disclosed the method yet. (Though they have previously had an official model that could query wolfram-alpha, so they're not strangers to that method)
But math olympiad questions have been beaten before by AlphaGeometry and a few other's using prolog or similar logic evaluation engines. And it works quite well. (Simply searching LLM prolog gives alot of results on Google and Google scholar)
If openai did it through brute forces text reasoning, its both impressive and frighteningly inefficient.
Even just normal algebra is something llms struggle with, hence using existing algebra solvers is faar more effective.
We should make insurance companies not allowed to negotiate special pricing for drugs and hospital expenses, and make everyone pay the same regardless if it goes through insurance, which insurer, or out-of-pocket.
Then the cost intensive flips. Insurer wants cheap healthcare and drugs so that they won't have to pay as much. This was part of what the original "affordable care act" tried to do, but was ultimately removed from the version that was passed.
It's also how insurance works by default almost everywhere except in the US.
The us spends more money on healthcare per capita than any European nation (including those with tax funded healthcare). Yet the very same drugs are cheap over here. The very same drugs Europe produce, put in airplanes and fly over to the US for the US market.
Are Europe just better at R&D then? Does Europe have more lax medication regulations? That is what your argument would suggest. But i somehow doubt that.
Looking at the share prices of the top medical industry companies in the US, from insurance to medicine production to private hospitals, it would seem there is plenty of margin going elsewhere for some reason.
Are we also ignoring that a lot of medical R&D is funded by grants and government investment? Its odd how the pharmaceutical companies are sooooo strained for money from the (partially already paid for) R&D that they have to take out a 600% margin on the product to cover it for decades after the drug has been on the market.
But it's clearly the famously harsh American bureaucracy that cripples the US market compared to Europe and Asia (the very same bureaucracy that created a self inflicted opioid crisis by being overly swayed by pharmaceutical lobbying)
I think you'll find if you spend a quick second googling, that Pharma is by and large the biggest spenders of money on R&D out of every major industry as a % of revenue.
And what you're failing to account for, is all of the failed research that they do that goes nowhere. Yes, they charge big margins on old drugs, however what do you think happens to all the money that gets spent on drug research that doesnt do anything?
Most medical research doesn't go up in smoke if it didn't lead to a production drug. And i don't argue R&D isn't a larger portion of what pharmaceutical companies do.
Although the statistics is old (2017), its interesting to compare the $2.2 billion total spent by all US healthcare companies listed (2017) compared to the $38 billion tax money spent by the NHS on research grants to medical research 2024. (80% of their budget of $48 billion)
Even assuming very generous increase of the medical research budget since 2017, makes N.H.S. research grants a significant portion of medical research that the expensive drugs supposedly cover?
And again, this is somehow only an issue in the US. Why does European and Asian pharmaceutical companies produce cheaper drugs and cutting edge medical research as a competitive rate, if the high cost of the drugs is to needed to keep them afloat?
The statistics you sent also seem to show non-us pharmaceutical companies regularly spent more % of their revenue on R&D. What gives?
---
And again, i didn't single-out pharmaceutical companies specifically. I blamed all of US medical industry, from the hospital to the insurance. The pharmaceutical companies seem to charge quite reasonable prices to the insurance companies. Odd how much more it costs to buy it without insurance though (the cost is to cover R&D right?.... right?)
---
I'll be even more spicy. If your goal it so save as many people as possible, and R&D truly is single-handedly the reason drugs and drugs and medical care is so expensive (which i think is blatantly incorrect); The US should just stop doing R&D entirely.
You could save vastly more lives with the drugs already in existence, which is prevented because the US healthcare system is so fundamentally broken.
But i don't think it would do squat because there are a lot of other things affecting the price of drugs and healthcare than that. Cutting out R&D not do squat on healthcare costs.
If the AI can only decently do JS and Python, then it can fully explain the observed disparity in opinion of its usefulness.
reply