I actually really like using an LLM for this, but only if I'm hunting for a phrase and I'm having trouble getting it to manifest. The LLM never gets it right, but it often shows me more the shape of what is wrong, and helps me to my creative end.

I try not to.

Well said. Maybe what we really need is regulatory oversight to standardize the way these things are used. Something to mitigate or stop abuse of the tools.

I've found if you are careful with your prompts, you can avoid the hallucinations. Moreso with Copilot, where experts (I hope) have made a good system prompt.

I do agree that it's often wrong. It demands constant vigilance, and I worry about companies pushing it without also pushing training.

I've used AI for exactly that. But I try to stay focused on the root of the problem (why I asked). So really I use the AI to find that gap, and let that gap be the key driver of what I should learn next.

Pretty handy to parse documentation, too, if you jump a lot between languages or platforms.

I'd go one later deeper, past addiction to money.

From what I've seen, modern AI models and the systems built with them could really add value to businesses. It's likely to be more profitable to sell these solutions to businesses, unless and until the compute required for LLMs becomes cheap.

Those three matching buildings towering over the skyline look incredible.

I find the idea of blockchain voting very interesting. There's a company trying to build it, but I can't remember their name.

I view the global Internet as an institution, and so long as it stands I think Cryptocurrencies will endure. So long as there are people supporting it.

I don't concern myself with what Trump cares about.

But I do like crypto, at least as an idea. I'm more interested in the function of the consensus mechanisms than I am about new chains.

Thank you for posting this for free so that I could read it.

When the Internet
goes dark
we will have much bigger problems than money.

This is really interesting. I always love to see these alternative compensation methods used in real business.

no u

This is dangerous rhetoric.

It sounds like the No True Scotsman fallacy, and I fear it will encourage us down a tragic path.

Do their crimes require them to die? A jury of their peers should decide. But we must not fall into the trap of dehumanizing those we see as evil.

I'm so confused. Is this satire?