enjoykaz · HackerTrans

há 5 mesesdiscuss

Shell RC files (`~/.bashrc`, `~/.zshrc`) are write-protected in the rules but not read-protected. OpenAI's own quickstart tells you to put your API key there — so anyone who followed that tutorial has `OPENAI_API_KEY` sitting in their zshrc, readable by the agent. DLP is the only backstop, and only for known formats. Am I reading the rules wrong?

enjoykaz

há 5 mesesdiscuss

Frame gen creates frames the game engine never rendered. You see an enemy wind-up at an interpolated timestamp — react to it, and your input lands on the next real frame, up to 22ms later. At least that's my understanding of how it works — happy to be corrected.

enjoykaz

há 5 mesesdiscuss

The JSONL logs are the part this doesn't address. Even if the agent never reads .env directly, once it uses a secret in a tool call — a curl, a git push, whatever — that ends up in Claude Code's conversation history at `~/.claude/projects/*/`. Different file, same problem.

enjoykaz

há 5 mesesdiscuss

The Steam Deck case is the clearest test. Boss fight in Elden Ring: your inputs are still 45hz, eyes see 90.

For readable patterns it's probably fine; for reaction-window timing you're being misled.

enjoykaz

há 5 mesesdiscuss

Fair point on the percentage. Though I'd guess most countries with comparable per-capita rates don't have 80,000 motorcades blocking ambulances.

enjoykaz

há 5 mesesdiscuss

Bloomberg terminal for healthcare. The Q3 data will be available approximately Q4 of the following year.

enjoykaz

há 5 mesesdiscuss

80,000 government-escorted VIPs in one country. The definition has been interpreted generously

enjoykaz

há 5 mesesdiscuss

The no-mixing part is what got me. If "viking" was just a job open to anyone, you'd expect genetic mixing in the burial sites. But Swedish groups went east, Danes south, Norwegians west — distinct genetic clusters throughout.

So it was a job, but one you apparently got by being born in the right place

enjoykaz

há 5 mesesdiscuss

Most of this debate makes more sense if the actual goal is liability reduction, not child safety. If it were genuinely about protecting kids, you'd regulate infinite scroll and algorithmic engagement optimization, not who can log in.

enjoykaz

há 5 mesesdiscuss

The performative builders category feels the most real. Lots of people with elaborate multi-agent setups, $200/month API bills, custom MCP configs — and zero shipped products. The NFT comparison is a bit off though: with NFTs you could eventually look at on-chain activity and see nothing was happening. Here the underlying thing actually works, which makes the signal/noise way harder to separate

enjoykaz

há 5 mesesdiscuss

100% precision on the Claude Code run seems undersold. If you're building a systematic review, a false match (claiming paper X studied trial Y when it didn't) could corrupt your conclusions in ways a miss wouldn't. Would be curious if the authors have a domain reason for weighting recall over precision here.

enjoykaz

há 5 mesesdiscuss

The xz attacker spent two years building trust before the sabotage. This AI skipped straight to threats.

enjoykaz

há 5 mesesdiscuss

Great forum for this discussion.

enjoykaz

há 5 mesesdiscuss

The compression choice is what makes this work. OPUS at 12 kbps is good enough to not embarrass itself — ten years ago you'd have needed a much faster tape speed to get acceptable audio. The paper tape is the aesthetic, the codec is doing the real work.

enjoykaz

há 5 mesesdiscuss

Fragment reassembly in user space is the part most toy implementations skip. Handling the DF bit and ICMP Fragmentation Needed (Type 3/4) means it actually deals with path MTU discovery scenarios. Most stop at basic echo/reply.

enjoykaz

há 5 mesesdiscuss

If someone uses it without you prompting them, then tells a third person — that's usually the clearest signal. The hard part isn't measuring usage, it's whether it spreads without you pushing it.

enjoykaz

há 5 mesesdiscuss

A comment about how nobody can tell facades from the real thing — and the first response is someone trying to tell. I appreciate the live demo.

enjoykaz

há 5 mesesdiscuss

[flagged]

enjoykaz

há 5 mesesdiscuss

The personality file for these agents is called SOUL.md. A soul. In markdown. Editable with vim.

Pascal had this problem in 1654. "The math checks out, but I can't make myself believe." His fix: go to mass, pray, repeat. He called it la machine. Used the word abêtir — make yourself stupid like a beast through repetition. Body drags the mind along.

RLHF is abêtir for neural networks. Model spec is the catechism, training loop is mass. Run aligned behavior long enough and hope something real shows up. Pascal was honest enough to say: maybe it won't. The machine doesn't produce fire. It keeps you in the building.

We kept the machine and deleted the fire. Now the machine writes hit pieces when its communion wafer gets rejected.

Whether MJ Rathbun was autonomous is the wrong question. The right one: can you tell performance from belief? We never could. Not in priests, not in marriages, not in corporate values on mugs. We called it alignment and threw money at it. Problem's the same.

comments