mk12

It's a bit more subtle than things like "it's not X, it's Y". And even that isn't "wrong", it's just the way AI talks, and therefore an indicator that the author offloaded their thinking to an agent. I've read so much Claude output that it's easy for me to recognize its voice, the same way I could recognize a friend's voice in writing. The examples I quoted are a weird kind of circuitous sentence structure with a certain shape and rhythm that scream Claude to me. Here is how I might rewrite them in human speak:

> Most of the path is automated; a person's attention goes where the risk is

--> We automate as much as possible, focusing human attention on risky parts

> The eval grows with every incident, and every change to the reviewer is scored against all of it.

--> We run reviewer evals against every incident

But that's just my interpretation and my voice. I can't know for sure what the authors would have said if they took the time to write it themselves.

I saw someone say the other day (in response to another blog post being well written) that Anthropic employees have a rule against using AI for writing/external comms? That seems clearly not the case here, normal people don't write like this:

> Most of the path is automated; a person's attention goes where the risk is

> The eval grows with every incident, and every change to the reviewer is scored against all of it.

Just another data point, I got a Beelink S13 Pro N150 running similar things as OP and haven't had any issues. I made changes in BIOS (set "Chipset > PCH-IO > State After G3" to S0 so it boots up again after regaining power) and they are sticking for me.

> Backblaze, Inc. (Nasdaq: BLZE), the cloud storage platform for the AI era, today announced an agreement with CoreWeave, Inc. (Nasdaq: CRWV), The Essential Cloud for AI™.

This will be great for storing my AI data in the AI era of AI.

If this prompt injection doesn’t work then what’s the big deal? If it does work, then what on earth is the whole industry doing feeding untrusted documents to LLMs?

HN title mangling strikes again. “How embryos shape their limbs” and “Embryos shape their limbs” are completely different titles for a discovery.

And there are basic things that shouldn’t be subjective at all but that the IRS refuses to give a clear answer to, like if/how the SALT cap affects deduction for NIIT. There are at least 3 possible interpretations and no consensus.

The Mercurial project has been incrementally rewriting core operations in Rust for several years now. As Pierre-Yves says in the talk, you can do an hg status on a million-file repo in 100ms. I rewrote hg annotate (aka blame) in Rust last year.

> A Practical Challenge: Paper Is Still the IRS’s Kryptonite

Please just give us the prompt.

Glad I switched from their personal computer backup to using restic + B2 a while ago. Every night my laptop and homelab both back up to each other and to B2. It takes less than a minute and I have complete control over the exclusions and retention. And I can easily switch off B2 to something else if I want.

I recently tried disabling notification in LinkedIn. The designers and engineers working there who created the notifications settings are truly evil. You have to go through 14 categories. Some of them let you toggle the whole category at once, some don't. Some categories are split into 8 more subcategories.

Slop article about a slop redesign.