If you write 3 bullet points and produce 500-pages of slop why would my AI summarise it back to the original 3 bullet points and not something else entirely?

It won't, and that's the joke. They will write three bullet points, but their AI will only focus on the first two and hallucinate two more to fill out the document. Your AI will ignore them completely and go off on some unrelated tangent based on the of the earlier hallucinations. Anthropic collects a fee from both of you and is the only real winner here.

This will never stop as it would require either the reference date to be changed or fir all dates in all saved spreadsheets to be off by one.

Or... A version number that tells excel which convention to use?

I suppose that would make copy and pasting formulas between spreadsheets very mildly error prone though, so it probably won't happen.

There already two date systems, IIRC for compatibility with Macs, I probably read it in the OldNewThing, but I can't finf it. Anyway, it's explained in https://support.microsoft.com/en-us/office/date-systems-in-e...

IIRC, it will never stop because now it has been "standardized". The OOXML specification explicitly mentions that the conversions of dates and their attributes (like whether a year is a leap one) has to follow the Excel implementation.

The US has traditionally solved this problem by having dozens of political entities that can compete (at least for elites) and, since the creation of the interstate highway system, the oppressed can flee.

In the UK open banking was essentially a response to GDPR this has allowed (to a limited extent) a variety of tools to be built on top of bank accounts that others would not have been.

That was actually the two Payment Services Directives: https://blog.finexer.com/guide-to-psd2-regulation-for-open-b...

Can someone explain to me what the difference is between a union in which everyone is a member and a government.

This might be the biggest benefit of AI coding. If I have a large legacy code base I can use AI to ask questions and find out where certain things are happening. This benefit is huge even if I choose not to vibe code anything. It ends up feeling a lot like the engineer that wrote the code is still with you or documented everything very well. In the real world there is a risk that documentation is wrong or that the engineer misremembers some detail so even the occasional hallucination is not a particularly big risk.

> This might be the biggest benefit of AI coding. If I have a large legacy code base I can use AI to ask questions and find out where certain things are happening. This benefit is huge even if I choose not to vibe code anything.

I definitely agree with this.

This.

Once you have revenue you have downside to protect. Pre-revenue the worst that can happen is that you have to start again knowing more than you did.

That would not have made it to the top of HN.

I would pay for YouTube premium again if I could switch off shorts.

I actually stopped paying for premium to make YouTube less appealing.

Another technique is to turn off watch history. As soon as you do that, you get a blank page on YouTube. It puts the decision making on the user to choose what to watch. I rarely get into the rage bait or shorts. It’s shown in search results and the sidebar. But at least, it’s not in the face when you open the website.