It was astounding just how quick it was, which really shows you how little of a damn they seem to give about their users. That's especially weird, given that they dogfood their own products.

Local backups are important, they're cheap, and often fast. They just shouldn't be the only kind of backup you do for data that is important to you.

opinion my own, yada yada yada.

In numerous cases, though, there are folks asking it to go interrogate some stuff they've set up MCPs for, and produce reports from it. If you do that it will give you a different answer every time, even from exactly the same input (because that's how LLMs work) and you just can't guarantee that any of them are accurate. It's a probabilistic layer, and the reports you need to generate need to be deterministic.

The problem is we're so accustomed to the deterministic nature of the large majority of the software we work with. The output is plausible, too, which only exasperates the problem. Folks just assume it's correct.

I'm not sure there's any good way to lay off large amounts of staff (besides not getting yourself into the situation in the first place where you have to)

On a site dedicated to commenting on articles? I think you have a misunderstanding of how HN works. People (hopefully) read the article and share uninformed^H informed opinions on the article.

That has always included critique of the way that the content is written.

In this case, very valid critique. I'm astounded you're somehow managing to read "A LOT" and not run into it regularly. At least we seem to be moving away from the absolutely awful "I'm a crazy frat bro" style of writing where it feels like half the action sentences should be appended with "because I'm crazy!" that was spreading far too far and wide (hopefully because it's hard to coax AI into that style.)

So e.g., don't use an LLM to call an API to gather data and produce a report on it, as that's feeding deterministic data through a "bullshit" layer, meaning you can't trust what comes out the other side. Instead use the LLM to help you write the code that will produce a deterministic output from deterministic data.

I've seen co-workers use LLMs to summarise deterministic data coming from APIs and have reports be wildly off the mark as often as they are accurate. Depending on what they're looking at that can have catastrophic risk.

#omnibox-ml-url-scoring-model

#omnibox-on-device-tail-suggestions

#optimization-guide-on-device-model

#text-safety-classifier

#prompt-api-for-gemini-nano

#writer-api-for-gemini-nano

#rewriter-api-for-gemini-nano

#proofreader-api-for-gemini-nano

#summarizer-api-for-gemini-nano

#on-device-model-litert-lm-backend

Then around gemini but not caught by the search for models: #skills (maybe? I think this is implied by "gemini in chrome"?)

edit: I don't see a carte blanch AI disabling option. As much as I dislike Mozilla's growing obsession with AI, at least they give me a top level option to disable all AI stuff. I only keep Chrome around for occasional testing reasons.

I've met so many people over the years who've worked in the industry at some stage (including myself), and not one of us has had a positive experience.

All that to say, it's almost certainly true that unless you've specifically heard otherwise about a hosting company, you're probably supporting assholes whoever you use.

Hawaii's one is at https://auctions.ehawaii.gov/dot/welcome.html. I had a boss when I lived in Hawaii that would bid on them, occasionally brought boxes of the random "sharps" (as they were called at the time) that had been confiscated from people by TSA. Some of it is somewhat surprising that people still had on them when they got to TSA, like high quality 18" long machetes.

They've done themselves no favours at all with their write up.

It does seem legitimate (I was able to use the PoC on a 24.04 instance), and seems like it should be a big deal, but the actual number of affected distributions seems way lower, and not even remotely as per their claim every distribution since 2017.

For example with Ubuntu, if I'm reading it right there's some impact in 16.04 (EOL), but then at least as per their analysis, only the vendor specific 6.17 kernels they ship that have it (e.g. linux-gcp, linux-oracle-6.7 etc.). That's a relatively new kernel version they started shipping recently, after it was released upstream last September.

I was thinking about fossil in the context of agentic workflows the other day, after seeing a co-worker go all in on sort of shifting themselves to a TPM workflow, using a locally hosted kanban board (inspired by OpenAI's Symphony).

It'd make things easier to have everything shoved into the repo, other than that everything is now shoved in the same repo being handled by the barely constrained chaos monkey that is an LLM coding agent. Locking things down gets hard if it's got access to the whole thing there.

After reading the linked article, and the comments here I still have zero clue about the game. It's a puzzle game involving sausages and a large fork does nothing to describe what kind of puzzles they might be.