hollownobody

I would use Sonnet instead of Opus because it's faster. Isn't it? It's a smaller model

Your wife doesn't let you use WeChat because you could find prostitutes? Sorry, this confused me. If you want to hire a prostitute, not having access to WeChat won't block you from that. Or is this more about avoiding temptation?

Are there any predictions you'd want to make? Not about AGI, but about an intermediate goalpost you think we won't reach in the next 5 years

720 tool calls? Amazing!

This reminds me of: https://en.m.wikipedia.org/wiki/Flying_Machines_Which_Do_Not...

"AI tools are bad", says company most suffering by the use of AI tools

Try Claude code, codex-cli, or gemini-cli. They are CLIs, as their name indicates. The Gemini one is free.

Try o3 please. Via UI.

He posted on Twitter about the coin and a few hours after the rug pull said that he hadn't done any research.

AFAIK, these models can't "look" at the plots they build. So it is necessary to send the screenshots, otherwise they will never notice the gorilla.

Try reading Qualityland, by Marc-Uwe Kling. It's SF (funny SF) about this exact premise.

One option is that people you talk to look for the stuff to discuss.

E.g. person A talks about bouldering with person B. Person A then continues with their day while person B googles bouldering terms.

A and B then receive both bouldering ads. Person By because of the googling and person A because of proximity to person B.

Person A then assumes the phone spies what they say, but it doesn't. Location is enough.

Done! Email subject is "HN DS Interviews".

Yes please!

For additional context, we are currently using problems which have test cases, and it feels as if it doesn't correctly reflect the day-to-day of a data scientist. In fact, many interviewers pass a candidate iff they pass the test cases.

I'd love more open-ended problems.

Not sure if this is not currently possible, but what about adding an LLM integration?

I work in DS for a large corp and we're currently working on replacing our hackerrank coding interview with something that lets us evaluate the daily life of a dev a bit more realistically. This includes letting them use LLMs and watching them do exploratory data analysis, instead of just coding.

I know this comment doesn't add much to the discussion, but why, god, why does the author feel the need to write completely in lowercase? It seems completely unnecessary. Impress with your words, not with a weird formatting.

I think you assume that there will only be one model. However, as we can see today, there can also be a variety of prices/qualities. Why wouldn't the tier 1 support people be also replaced, but by cheaper models?

I disagree. LLMs are creating new knowledge when they are being used. A conversation with an LLM can carry a lot of signal in the user's input. I routinely send whole snippets of code, alongside an explanation/question as input.

An European country perhaps? If you get a visa for q few years in Germany, for example, then you can apply for permanent residence. There are freelance visas.