__0x01

Agents are very good at making us think the tower is rising, when in fact it is falling beneath our feet.

Thanks for this. I especially enjoyed the section on word counting software and word linking games.

From the companion paper:

The argument relies crucially on ideas that may, at least in retrospect, be attributed to Ellenberg-Venkatesh, Golod-Shafarevich, and Hajir-Maire-Ramakrishna.

Can someone please elaborate on this?

A problem with this is that in order to confirm the findings, you’ll need an expert human. But generally expert humans are busy doing other things.

The article suggests using LLMs to identify and fix UB. However as per the above, I think the issue is that we need more expert humans.

LLM generated code will eventually contain UB.

EDIT: added "eventually"

Correctness: 1,778 test cases from the official jsonata-js test suite + 2,107 integration tests in the production wrapper.

The AI generated code can still introduce subtle bugs that lead to incorrect behaviour.

One example of this is the introduction of functions into the codebase (by AI) that have bugs but no corresponding tests.

EDIT: correct quotation characters

Can a datacenter practicably upgrade to the next generation of GPU every year?

What was the interesting problem that Swiss watchmakers could have focused on, instead of pursuing brand?

Please can you elaborate? We are more intersted in "the end" in what sense?

Is "AI code review" a correct term?

A code review requires reasoning and understanding, things that to my knowledge a generative model cannot do.

Surely the most an AI code review ever could be is something that looks like a code review.

This is nothing but speculation

Did you read the paper?

I also worry about a centralised service having access to confidential and private plaintext files of millions of users.

Often when I look closely at the output of LLM generated code, I see repetition, redundant logic and deeply hidden bugs.

Notwithstanding the above, to my understanding LLM services are currently being sold below cost.

If all of the above is true, at some point the degredation of quality in codebases that use these tools will be too expensive to ignore.

These LLM tools appear to have an unprecedented amount of access to the file systems of their users.

Is this correct, and if so do we need to be concerned about user privacy and security?

Please can you provide a source for the above?

My understanding was that atherosclerotic plaques are comprised of cholesterol or fatty deposits [1] and that these can lead to CVD.

The fat mechanism I understand, but what is the mechanism for sugar in CVD?

[1] https://www.heart.org/en/health-topics/cholesterol/about-cho...

Engineering Room, panning over a bunch of hot Blackwells

"I can't change the laws of physics!"

The monster babbleth no more, sire.

Is cost a risk here? I'm assuming that sometime in the future the price for vibe coding/engineering will go up significantly.

I have noticed that many of the most used and loved projects use C/C++.

With all of the unpopular press that they get, why has history often proven C/C++ to be the right choice time and time again?

Is this cheaper to run than the gas equivalent?

What LLM does Phoenix.new use?

I thought this was going to be an essay on the impact of English, Math and Programming on humanity.

I would place English (or all spoken/written languages in general) first, Math (as discovered) second, and programming languages last.

honest-broker is the only Substack publication I have found to consistently hold my attention. Reading Gioia's writing feels like eating nutritious food.

I am most likely ignorant, however. Are there any other Substack publications of the same quality?

This is the most prescient point I have read hitherto regarding LLMs. We are fashioning for ourselves gods of wood and stone.

Is this designed to be run in production?

Please could someone explain, very simply, what the training data was composed of?

"October 2008

The economic situation is apparently so grim that some experts fear we may be in for a stretch as bad as the mid seventies.

When Microsoft and Apple were founded." [1]

"A programmer can sit down in front of a computer and create wealth. A good piece of software is, in itself, a valuable thing." [2]

These words give me hope.

[1] https://paulgraham.com/badeconomy.html [2] https://paulgraham.com/wealth.html

Error level logging can exist with a metrics focused approach.

What books can I read to reason like this?

EDIT: shortened sentence

My final draft is often completely different to my first draft, which I wouldn’t want to ship.

Most of the time I only start to see the real shape of the solution after a few days of work.

Those iterations each day have value as they propel me toward a solution. But it’s more of an internal type of shipping. I’m shipping to myself.