HN user
The opinions expressed may not reflect the views of my employers, my business, or myself.
The issue associated with whataboutism is that it is a deflection strategy. It aims to steer the focus away from the issue at hand. Commonly regarded as intentional manipulation or lack of adherence to social conversation norms.
> The required high degree of connectivity might be attained with the help of microminiature radio transmitters and receivers.
Hahaha, this is straight out of 60s-70s sci-fi, where their best futuristic interfaces were smaller CRT screens / flashy keys, etc.
Actually that is not far off from how high bandwidth interconnects work, but through other media.
Heat exchange is used instead of refrigerating the coolant. Makes sense. How do they manage the indoor climate for the humans working there though? Eventually everything will be at 45C in the building, will it not?
Requesting human attention demands demonstrating human effort[1]. Slop at its worst: obscuring actual insight. The main one being that there are no exponential growth curves, no hockey-sticks, they're all logistic s-curves at most. Sure, there is a lot happening right now, but although technologies such as trains reshaped society, society did not come to only consist of railways.
The FFT is essentially just a matrix multiplication, or two. No need for fancy conversions. Just a huge amount of training data and a very large array.
The power needs of a house of course vary with many parameters. Our 100 square meter house in Nordic climate draws 1.2 kW on average over a year including charging the car. What are common numbers elsewhere?
Something like that.
This was of course dependent on yolo mode, but automatic approval has also been pulling stunts like this. A recent example is data that was purposely kept away from Codex in a folder far far away. When it found a single reference it just went for the data when having an issue. Lesson learned, keep essential data and Codex separated on different machines. Codex remote ssh actually helps here.
The closest we've ever been!
”Religious suffering is, at one and the same time, the expression of real suffering and a protest against real suffering.
Religion is the sigh of the oppressed creature, the heart of a heartless world, and the soul of soulless conditions.
It is the opium of the people.”
Some are on copium, some on hopium. The gods change names; the need for relief remains.
Keen observation. Maybe automation will come for the AI as well?
The people were longing for utopia, just not the same utopia.
Another argument for less human-like AI then, I guess.
I'm starting to long for the age after AI. When the generative euphoria has settled and all outputs are formally verified based on exquisite architectures and standards.
I see your point. Many of my prompts for reasoning ends with: No code. Planning mode is sort of the workaround for this specific situation. Sometimes it is useful for the AI agent just to think. It looks like I need a screwdriver in addition to the aforementioned hammer, a pozidriv screwdriver to be precise.
To be fair, there is likely not much training data on the difficult conversations you need to handle in a senior position, pushback being one of them. The trouble for the agents is that it is post hoc, to explain themselves, rationalising rather than ”help me understand” beforehand.
I think that it is a fair perspective to allow role play, and it's useful too, when explicit. Does not really make sense for AI to cosplay human all the time though.
Maybe we should change that? Of course symbolic AI was the holy grail until statistical AI came in and swept the floor. Maybe something else though.
The other day Codex on Mac gained the ability to control the UI. Will it close itself if instructed though? Maybe test that and make a benchmark. Closebench.
Asking for code to manipulate the AST is another route. In python it can do absolute magic.
Interesting. When I code, I want a boring tool that just does the work. A hammer. I think we agree on that the tool should complete the assignment reliably, without skipping parts or turning an entirely implementable task into a discussion though.
That is probably the next step, and in practice it is much of what sub-agents already provide: a kind of tabula rasa. Context is not always an advantage. Sometimes it becomes the problem.
In long editing sessions with multiple iterations, the context can accumulate stale information, and that actively hurts model performance. Compaction is one way to deal with that. It strips out material that should be re-read from disk instead of being carried forward.
A concrete example is iterative file editing with Codex. I rewrite parts of a file so they actually work and match the project’s style. Then Codex changes the code back to the version still sitting in its context. It does not stop to consider that, if an external edit was made, that edit is probably important.
Agreed. I am asking for something beyond the current state of the art. My guess is that stronger RL on the model side, together with better harness support, will eventually make it possible. However, it's the part about framing the failure to do complete a task as a communication mishap that really makes me go awry.
It's not necessarily "ignoring" instructions, it's the ironic effect of mentioning something not to focus on, which produces focus on said thing. The classic version is: "For the next minute, try not to think about a pink elephant. You can think about anything else you like, just not a pink elephant."
Agreed. We should not be anthropomorphising LLMs or having them mimic humans.
Codex in this case. I didn't even think about mentioning it. I'll update the post if it's actually relevant. Which I guess it is.
EDIT: It's specifically GPT-5.4 High in the Codex harness.
On point. I'm more interested in what comes after LLMs/AI/AI-agents, what the next leap is.
My point was narrower than suggested. If Mythos is in fact a security risk, then the NSA is one of the actors most likely to already understand that. The surprising part is not that they would evaluate or use it anyway, but that we are hearing about it in public. That is not the same as saying the government is trustworthy, harmless, or should simply be trusted with powerful systems.
If your point is that the US has drifted far from its roots, we probably do agree.
That is expected. What is not expected is us knowing about it. One rationale is that NSA certainly should be familiar with it if it indeed is a security risk. Nothing to see here.
This pushed me to try pi.dev with a remote ollama instance running gemma4:26b on a 24GB RTX 3090. It is not perfect, but it's quite okay-ish with 128k context length. It's not a model primarily for coding I guess, but IIWIW. Thanks!