wswope

Fair play; I got too personal here. Apologies.

My approach to this has been a NixOS host with the harness running in a rootless podman sidecar.

The host has squid configured with a self-signed CA and networking rules to route all host traffic to the intercepting proxy, so I have a tight firewall and full auditability.

Then there’s a python rpc daemon running on the host with a set of whitelisted commands, read-only for pulling logs and diagnostics.

By default, the agent runs in a split pane tmux session with a host shell on the left and the chat interface on the right. The rpc whitelist includes the proper `tmux capture-pane` invocation to pull from the host shell, so I can easily let it see what I’m doing if I want it to help debug something.

I’m using pi as my harness and have custom extensions that give Yes/No confirmation gates for any writes the agent makes and that pass all bash commands/file writes to a deepseek subagent for review.

Still early days, but as someone with a similarly paranoid mindset around running LLMs securely, I think the future is promising and we’ll see some new “best practices” and related tooling popping up shortly.

That’s been the pitch of Slate Auto, which is supposed to release its standard model at the end of this year.

https://www.slate.auto/en

Unfortunately, they’ve been suspiciously quiet since their initial announcement, but fingers crossed. Might actually have a viable new car option if they’re successful.

Not OP, but having explored the field a good bit, Openrouter + pi harness in a devcontainer work great as a sane starting point.

Highly recommend as a clean way to try out the upstart models.

See also: privacy.com

(Virtual card provider that generates cards as a free-to-the-user service. They make their money from a cut of the standard transaction fees. Cards are locked to a single merchant and it’s easy to configure limits.)

This is a horrendously bad-faith take. You know full well it’s *not* just a one-off $200 issue: they treat customers like this at scale.

Don’t pretend this is an isolated matter, or that CS/billing is the only arena where Anthropic has such systemic issues.

I don’t know you, but your response honestly reads like it’s coming from someone wrestling with their own moral compromises. If so, please take a good hard look in the mirror. (E: yep — https://news.ycombinator.com/item?id=47953576)

What people typically mean by the GP statement is that the “thinking” mode of these models is loosely analogous to what humans do: a bit of a retrograde reconstruction of how we arrived at a gestalt conclusion that sounds good, but may not accurately reflect the real logic at play.

IME you can see this more easily with less-polished models like Deepseek 3.X, where the reasoning in the thinking traces occasionally contradicts or has zero bearing on the non-thinking output.

They were shirtless, unarmed and carried no visible communications equipment. They also appeared to have no idea what had just hit them, or that the U.S. military was weighing whether to finish them off.

"The video follows them for about an hour as they tried to flip the boat back over. They couldn't do it," one source said.

https://www.reuters.com/world/americas/video-shows-final-con...

Obligatory warning that PSL is led by full-blown tankies, and e.g., have an explicit pro-North Korea party line.

That said, the membership base is pointedly heterogeneous in belief and they’re the most active alt-left element in DC.

Make of this whatever you will; just offering you this as background for an informed decision.

Ah jeez, I was missing that detail. Yeah that’s messed up.

I skimmed by the “what gets sent” table and thought the bash telemetry was gated by the prompt-related opt-in behavior. Thanks for the correction!

Cmon now — I’m a rabid privacy nut but this is unfair given the context of:

Prompt telemetry is opt-in and off by default. The hook asks once; if you don't answer, session-end cleanup marks it as disabled. We don't collect prompt text unless you explicitly say yes.

The UUID part is just one accessory layer, and something plenty of other players in the ecosystem don’t bother to stick to.

Feels like actually bothering to ask users for consent is what got them burned here, when I’d say it’s at least an improvement that they’re asking at all. Many products don’t, and users never bother to turn it off because they don’t know and don’t care.

I think this whole UX is deeply misguided but at least has plausibly benevolent intent.

I'm sorry YT isn't working well for y'all. That said, expecting some grunt employee like me to feel personally responsible (or even "ashamed"??) is ridiculous.

The point of a company is what it does: The money in your paycheck comes from anti-competitive behavior, denying accountability and customer support to your users, and yes, even the enshittification of Youtube.

Deny it all you want, but you have every opportunity to walk away and do something better for society with your life if you so choose.

Exactly the sort of user hostility I’d expect from a google employee. Shameful.

Maybe if practical tools such as like-ratios were in place users could sort good content from bad.

Internet or not, I post under my real name on here, and I fully stand by my words. Anything I say on here, I’m 100% willing to say to someone’s face. We can link up for coffee or a beer next time I’m in CA if you’d like, and I’ll prove it.

The root comment is an aggressive affront to the audience’s collective intelligence. You’re in full “rules for thee; not for me” territory, and undermining your own site guidelines if you wanna let the root comment stand unchecked but go after the rightful callouts, in my book.

It’s not about bigco at all in my eyes.

At the end of the day, if you want intellectual curiosity and openness, bad-faith dishonesty needs to be weeded out; thought-provoking and honest conversation should be promoted, regardless of where the contributor is employed.

The problem isn’t working for Microsoft. The problem is dishonesty.

You’re treating the root comment with kid gloves because it’s from a Microsoft employee. Please don’t do that.

The impulse to hit back against what is perceived as a "transparently dishonest corporate actor" is natural and human.

Honest question: If we agree that the transparent dishonesty and the lynch mob behavior are both undesirable, how do you think the two should be balanced in operative terms?

I don’t want to put words in your mouth — but are you saying you won’t allow direct pushback to dishonest corporate actors??

My view is that healthy discourse requires balance and proportionality: flagrant dishonesty, as is the case here, should license a proportional degree of pushback.

I don’t agree at all that “nobody believes this” is quite the personal attack you’re making it out to be, but I don’t care to debate that at length either.

GP did not personally attack or denigrate the person they were replying to.

As the dozens of other comments show, the overwhelming majority of us do not believe the root commentors claims, and this PM quite objectively does not have the leverage and authority to back their claim that they won’t let this happen again.

It’s hard not to read your conception of “trying for something different” as granting undue credulity to a transparently dishonest corporate actor.

Sure looks to me like this whole case is Claude Code chasing its own tail, failing to debug, and offering to instead generate a bug report for the user when it can't figure out a better way forward.

Maybe even submitting the bug report "agentically" without user input, if it's running on host without guardrails (pure speculation).

E: It's a runaway bot lol https://github.com/anthropics/claude-code/issues/40701#issue...

I haven’t had a Windows box in about 8 years, but even back then all the big names had consistently better performance on Linux.

Usually about a 10-20% fps improvement for my usual fare in those days: League, Overwatch, Civ5, Minecraft, Crusader Kings, Factorio, etc. Try it for yourself and see what you get.

They closed it (which is fine) but there is no offline migrate alternative.

If it helps, you can copy the history table from one db to another in 3 simple lines of sqlite. I’ve done it myself a few times with zero fuss.

What stability, reliability, and performance problems are you hitting with Atuin?

I posted a longer comment upthread, but I’ve been self-hosting and running an old version for over two years now, and haven’t had any recurring problems on those fronts. It’s pretty damn stable software and everything they’ve been doing lately has just been extra features and gloss.

Not to detract from your point, but I decided Atuin was feature complete enough for my liking over two years ago, and have been running the last v17 release ever since with zero problems.

You can still keep everything you like about the tool without issue, and to the devs’ credit, the sync server is some of the easiest software to self-host possible.

Yeah, but this happens all the time.

>95% of the time, the fraudsters get off scot-free. Look at Dan Ariely: Caught red-handed faking data in Excel using the stupidest approach imaginable, and outed as a sex pest in the Epstein files. Duke is still giving him their full backing.

It’s easy to find fraud, but what’s the point if our institutions have rotten all the way through and don’t care, even when there’s a smoking gun?

The missing context is the title.

Just throwing these out there for people looking for options: Check out Piefed for a Reddit-like, and lobste.rs for an HN-like.

The fediverse might not be booming with activity to the same degree as the big names, but there’s plenty out there and it has a pleasant small-forum vibe that went extinct on the bigger sites a long time ago.

Can’t make a man understand something when his paycheck depends on him not understanding it.

Dan sold out his moral compass.

I’m plenty calm. There’s just nothing to debate here: the blog post and repo are a conscious, deliberate, and egregious misrepresentation of fact.

I would absolutely say exactly the same things to the author’s face as I’m saying right now. I would never work for a company that condones this in a million years, as a matter of principle.

Is it really worth it?

Unequivocally yes.

Fraud is fraud, and if your first instinct is to defend it in this manner, check yourself in the mirror.

Speaking as the user you responded to in the first link: I’m sure you get flak from all corners, but I (and presumably many others like me) take zero issue with your philosophy or design goals for the site. I’m with you all the way when you say you don’t want to HN to turn into a current affairs site.

The problem IMO is the current sensitivity of the filtering. You said upstream that a high enough number of upvotes could outweigh flags — but that CECOT thread was one of the top 5-10 posts of the month despite having been taken down for many hours overnight. If that wasn’t a thread where the upvotes should be a strong enough signal to override the flags, I dunno what is.

We’re asking for subtle knob tuning, not revolution.

I run an OSRS market analysis/flipping site, and have been keeping an eye on the effects.

The short answer is that there hasn't been a ton of movement across the market at large, but since Saturday, bonds have been swinging up towards the all-time high they set last December. Can't say for certain that that movement is tied to VZ though.