I usually listen to what doctors have to say, and trust their judgment. Especially since I have next to no formal education in bio or medicine.

I did resort to "self-diagnosis" once, though: for R-CPD (no-burp syndrome). Most doctors simply don't know this condition exists. I brought it up with my GP (whose referral I needed to see a specialist in Ontario), prepared for some pushback, but was pleasantly surprised that he did some brief research and then helped me forward.

Then again, I may have been lucky. I still hear stories of doctors who are less than cooperative when patients bring them "self-diagnoses", even for something as open-and-shut as R-CPD. But if most of those "diagnoses" are BS, then I can empathize with the docs, too.

Head to either side of the street. (one or the other); Plant trees on either side of the street. (both)

It's yet to be found. (not found); Yet it's found. (found)

They're all but forgotten. (nearly); They're known but to God. (only)

Does "all but" always mean "nearly"? I've been conversing in English for 20 years now but this one eluded me.

I guess I am just unable to parse the words to piece together such a meaning. In my mind, there are two ways to interpret it: "this subject may be all kinds of things, but one thing it is not" (all but = everything except), or the opposite "this subject may be all kinds of things, but one thing it is for sure". And neither of these means "nearly" :(

Yes. I think the idea is that on the path of events or actions between the starting point and the target, you have already completed every single one, and all that remains now is the final step before you can call it done.

You've walked all but the last step of the journey.

Thank you, that makes sense!

The outcome of the meeting was negative (bad); the outcome of the cancer screening was negative (good)

That one doesn’t belong - if the meeting was deciding the death sentence, negative could be good.

This resonates, but also, I'm not convinced anyone can truly understand the workings of any society, or market economy, or what have you.

Who's in control? Everyone is, to some extent. And no one is: when you're hungry for food, are "you" in control of that? You can consciously repress your impulses to go eat something, but your mind didn't create those impulses.

Human societies develop impulses and minds of their own, emerging (weakly) from the impulses and minds that comprise them, and they make decisions in mysterious ways.

Of course, it sure is nice when we can come up with a compelling story for the motivations behind something. Easier said than done…

As someone not super familiar with bankruptcy procedure: why do they auction off exclusive access to the data? Couldn't they make more money by, say, taking the top N offers, where N is "small" (e.g., ≤5)? Or does that depress the price too much?

In theory, a buyer could always execute that strategy after the sale, so it's essentially "priced in". By keeping the auction exclusive, the seller also keeps the possibility that the data is worth more to someone as an exclusive asset.

That's what thinking output is for, right? Mixing random tokens that live roughly in the same semantic realm, throwing them at the wall, and seeing what sticks? Hopefully, this backticks concern didn't stick.

I thought this was ironic because there's no good way to actually restrict which commands the AI runs save for sandboxing; and here it was expressing concern about OS command injection in the git commit message string argument to git. How to not use the equivalent of what subprocess calls (shell=True) which does exec code in backticks?

Would single quotes solve this

Which types of documents have this particular AI vulnerability?

At least for Claude Code, putting "run /simplify at the end" in an "implement the plan" skill helps a little. It still often leaves new code in bizarre places, and/or with bad/alien-sounding names and comments.

Yeah I would hold that models don’t know how to simplify because most rl/benchmarks doesn’t penalize complexity

Sloppy, non-privacy-preserving "analytics" set up by some communications intern ≠ malice. They can mean well and still do a poor job. True privacy on the internet is notoriously hard.

Sorry, no. Code doesn't magically go from the intern's workstation to live. If the chain of incompetence is that long, then the most generous read is that they don't care

Could've been as simple as installing and configuring some off-the-shelf "analytics" and "cookie banner" plugins for their content management system. Unfortunately, the chain of incompetence is that long, because these orgs do not attract talent that understands these things.

I don't think you can handwave responsibility like that, even if your theoretical were true it's still the responsibility of the org.

I suspect many competent devs in the industry would find it sensible if Anthropic used their products as light-touch "assistants" sometimes. But yeah, it wouldn't fit the outside narrative that's formed and conveniently propped up valuations.