polotics

At the end of the first paragraph there is this sentence:

"Training support is a hypothesis consistent with these results, not something this study proves."

I understand it, but I find the wording very unnatural. To improve readability I would have written it in the active form:

"We cannot prove training explains this result, but this fits the data best."

OK so there is a better angle to look at all this, there really is:

https://www.researchgate.net/publication/357868645_Edgar_All...

This is so weird, because using a script to invoke the agy cli is clearly using a "third party tool to access the service" !

What is preventing anyone from automating on top of the normal google Antigravity CLI?

I disagree.

This trolling and misdirection are not in the movie. Also, both are clearly effective, and a discourse that warps economic impact on US midterms voters within identities, reduces the chance of exposing the bad consequences and risks of power-politics. So this is quite well timed, and the outrageous quotient is I think also well calibrated.

Charles Chaplin understood it, so please y'all catch up ASAP.

More detail on the Patrick Boyle video where the actual picture of Bessent's note that included that purchase is mentioned. Also the fact that these are specifically French bonds that were dumped with no warning. This is dropping the mask of Western solidarity I think Xi is over the moon

well in my experience mostly you get either:

- stupidity

- stupidity and malice

I'm still very impressed by qubes, and glad I'm not such a target that I feel I need the level of opsec it affords (on all my laptops).

Maybe someday AI-assisted killchains will be so widespread that Qubes is the minimal level for the (few?) still-local users of compute.

I would not have copied anything from dom0 to any another qube, the impact is low.

please refrain from referring to countries as if they were individuals

The Savior fallacy. With power more concentrated, the ability to avoid being impacted for those in power rises and rises and rises. How are incentives then aligned? I'm sure that having 10 million governments would be much better than having one to solve the issue.

Make your own benchmark on your own work, keep them to yourself. Try one typical not completely unambiguous spec document like you're likely to have seen. See if the AI asks the right questions, and how it navigates its unknown unknowns.

I kind of agree that the post is a bit too far away from the trenches to be able to claim it will reveal "software engineering reality".

From where I snipe, I see a big divide between those...

1) that try to surrender to AI, aiming to fully replace value-added intellectual effort and often also to augment enterprise value-mask slop busywork...

...they fail, and succeed, and the collective suffers.

2) those that ride AI to get more challenged, more feedback if any kind, to tread further but with attention to the right details

...they succeed

obesity and food-supply are not in a linear relationship, pumping out some high sucrose corn syrup is one thing, having all the nutrients the population needs to stay healthy is another

This is hilarious, and so true!

Wow, that's a different perspective. I should read more of your posts. For me the fact that 30% of men and 20% of women are not having intimacy is quite scary. I guess a revised edition of the book sextinction is in order.

Well, also screwdrivers can do that...

if you actually do use AI to do some proper research on a fields you don't really know that well, looking for peer reviewed papers, actually figuring more things out thanks to AI, then I don't think brain damage will happen.

about your blog post: You jump from a quote about llms hallucinating articles references in students papers, to the statement "if AI does the thinking for you". since when is inventing falsehoods thinking?

if your agent's software delivers real world value to some people, that's motivation enough I find

I find that actually asking the LLM what they've done and why, and pointing out everything that's not quite right, is a good way to avoid that kind of overload. if you're actually managing to have fifteen simultaneous work trees all churning with agents, and all adding real value, on properly abstracted code not mindless copy-pasta or data filling, that is kind of a humble brag anyway, so write a blog post about it!

Did you really mean to write really: "Online girlfriends are cool"

In my humble opinion, they are the epitome of uncool.

Can you define your use of the word 'smartest' here just in case some of us don't quite know what you mean?

Have we really though?

This is not a silly question, because it's kind of silly:

Has an AI ever actually gone through all of the steps necessary to become a chess master? Go to competitions, raise in the ranks, take a plane to the masters, organize support and all the real life details necessary? Playing chess is just the small formal part here, and all of the obvious things humans obviously do, well...

this is software development

So exactly true! With ideas people unbridled, the painting-yourself-in-a-corner tendency that was always there has gone exponential, add to it the not-completely-aligned incentives, and the oh-so-clever impressive writer of long prose LLMs just end up churning. The real value add may end up being the lesson in honesty and humility to us all...

Yep and I just ran a simple enterprisey "ambiguity" bench on the big three (US) model providers: same ambiguous initial-prompt with same clarifications and pushback prompt sequence afterwards.

The edge of correct/better when facing ambiguity is very fuzzy, all models from the past 6 month or so have similar random ways of spinning between too-literal avenues and oddly misplaced misled fixations. Taking the right initiatives in face of uncertainty is definitely AGI, and its not there, and perceptrons + attention layers just ain't got what it takes no matter how hard you push.

This quote from the post makes me sad for all the researchers and all the math and the scaling and the investing that I believe was done in good faith in pursuit of pure science and betterment for us all:

"AI is designed to make you dependent on it"

This is throwing a lot of agentic opacity bathwater and not really looking what else's in there.

Hi. This is very interesting, could you link to the research? There is a dearth of proper research studies that A/B test what approach is best in terms of harness structure based on repeatable benchmark data with relevant sample uses-cases.

can you point us towards the specific tells you're seeing in the article. I find the writing maybe LLM-cheerfull yes, but don't see classic phrase patterns.

René Girard's theories on mimetic desire come to mind, even though I sometimes thing his thought was not as universal as some assume.

Maybe he was just describing a sickness of self-domesticated humans, or even Neoteny (babies need attention more...) finding a way to evolutionary adaptation but with large side-effects and massive selective wastage.

Or maybe resource constraints and competitive pressure within some societies make it a rational strategy to accumulate pictures of a 3 weeks' restful holiday in a one-week selfie-stick scramble.

I am steganographically ensuring that my human writing is instantly recognizable by always leaving a few typos and spelling mistakes an llm would never commit. also long sentences. This kind of reminds me of Dune; you have to make sure to not step in rhythm, or the worms will come.

If I may quote you:

"Reasoning from a sibling product generates confident, plausible, wrong answers. These each burned real time before collapsing."

May I ask what exactly did you intend to mean when you wrote this sentence? it seems to me it's a bit all over the place, not sure what the purpose was in writing that one. It and many others here seem to be indeed burning some of my real time before collapsing into that question of: what's the precise intent.