> One command. Everything configured. Nothing to research.
Being talked at by someone's AI copypasta feels like being in the Truman Show.
"Why don't you let me fix you some of this new Mococo drink? All natural cocoa beans from the upper slopes of Mount Nicaragua, no artificial sweeteners!"
"What the hell are you talking about? Who're you talking to?"
What you're suggesting seems to go implausibly far beyond what the paper says.
RL post-training alters the parameters of the transformer, while your f(manifold) idea seems to suggest that a new layer on top would suffice, no need to alter the transformer itself at all.
It would be extremely handy if that were so, but I'm guessing it isn't, or it would be the prevailing approach.
A while ago a lot of the discussion about overparameterization was about explaining "double descent", the observation that test error doesn't descend monotonically and actually hits a local maximum around the point where the model has just enough parameters to interpolate the data. My favorite article about double descent looks at this in terms of splines [1]. If I can try to summarize that article: when you are designing a parametrized model to fit to data, you have a choice. You can either:
1. Avoid overparameterization by design. Manually create or choose a space of functions that has limited degrees of freedom by construction.
2. Accept overparameterization and regularize.
The latter tends to be more robust, because of the bitter lesson. It's not practical to manually design an ideal, on-demand, just-right limited-parameter model for every dataset we are presented with. The best way to approach that ideal, it turns out, is really to just let the computer figure it out via regularized optimization over an overparameterized space.
Statisticians started moving in favor of overparameterization long before deep learning got off the ground. This trend dates back at least to the machine learning bible, Elements of Statistical Learning (2001).
This app is a great example of what AI does to your brain. No one making their own choices in the app design would make each question need three clicks.
This weird trend reached an apex in a Feb 2026 OpenAI blog post [1], recently on the front page [2], which describes the process for building... something... written 100% by agents.
There is no description of what the thing is, no indication of what value it provides its users. The closest it gets is "the product has been used by hundreds of users internally, including daily internal power users".
But the fact that the thing has a million lines of code is repeated twice in the first few hundred words.
We've known for decades that output metrics like LOC/day are very bad measures of real productivity in software. But they seem to be back in vogue in the age of AI, because AI is so good at maxing these useless metrics, and we need to show how impressive our AI is and how impressive our usage of AI is.
This post and TFA have a common issue: no one seems to have a clear, compellingly evidenced account of basic questions about the collection and its history under consignment:
1. What exactly was in the collection?
2. What happened to the collection after it was consigned: which sets were sold, which were stolen or lost, which were moved to off-site storage, etc.?
3. How much money did the original franchise owner owe the consigner for the sets sold?
The peripheral claims about e.g. police malfeasance are disturbing, but without this basic evidence about the substance of the matter, I don't know if it's a great idea for an online mob to take sides.
What I keep wondering is, what would have happened without the AI? Would they have just ignored your request?
In my neck of the woods it's fairly common that when a person doesn't know how to help you they just don't reply, instead of saying "I don't know how to help, sorry". AI-generated responses seem like the evolution of this attitude that one must either ignore or respond in a (superficially) helpful manner.
Yes, this and every internet forum will still be doing this two years hence. Your life will be better if you take to heart this famous passage from Nietzsche:
I do not want to wage war against what is ugly. I do not want to accuse; I do not even want to accuse those who accuse. Looking away shall be my only negation.
I think they're saying it has qualitatively different capabilities that make certain kinds of security work more worth pursuing with the model, not that the model of human-AI interaction has changed.
You're right that they're using a harness like everyone else. The general idea of giving the model a harness is not going to change. I mean even humans need harnesses to accomplish some things.
Being talked at by someone's AI copypasta feels like being in the Truman Show.
"Why don't you let me fix you some of this new Mococo drink? All natural cocoa beans from the upper slopes of Mount Nicaragua, no artificial sweeteners!"
"What the hell are you talking about? Who're you talking to?"