Yes, I like malaise... even something like a "feeling of powerlessness" is often a result of that malaise, not the cause of it. Affordability is one of the things we can more objectively say is more perception than reality, but the same can still be true of subjective feelings

What's a project you fantasize about, but you also think it's too ambitious (or perhaps impossible to actually achieve outside of the dreamstate of imagination)?

Like right now an LLM can probably translate something like "the kitchen should have a good flow between functional areas" into a series of questions (eg "how far is the sink from the dishwasher")

A more useful LLM-assisted process (at least right now) would be one where the LLM can translate rules (formal and informal) into constraints that can be fed into a more deterministic system. Or maybe a layout that can be queried, so the LLM can probe and react.

How does one get a minneapolis.mn.us domain?

I still have one that boots... it has a nice physical design, still pretty unique. The keyboard and trackpad were impossibly terrible though. But the hybrid display, the integrated handle, and the stability when it's open are all great.

That an LLM can engage with our distinct humanity is also clearly frightening. But I think technology is more likely to magnify what it touches than it is to devalue or replace it

Reminds me of this article about boys and school, which had all these likes despite its complete disinterest in boys or any attempt at empathy with those boys… time.com/7335723/auto...

In defense of both bits of advice, it would be very convenient if they were true

Is a lot of the shipping in Europe done by ships registered/staffed from outside of Europe?

It never really occurred to me that Europe doesn't have much rail freight. How did it end up like that?

Please do leave notes here!

I'd definitely be open to trying something beside zod, but this seems almost exactly equivalent?

The features I wish zod had: schema-from-type (for lightweight validation), maybe better annotations (so I could define all my CRUD in schemas). Maybe that's it?

It reminds me of concerns about “appropriation” - there’s no legal concern, and the underlying notion of ownership has no relation to IP. They feel the LLM is appropriating what belongs to humans.

In their defense, those same people are also genuinely bothered by plagiarism, blogs that clone content, etc. Like they literally complain about it, which is a non-LLM IP/replacement concern. They don’t care about IP law, but they do have an intuitive sense of intellectual theft.

Btw the word that’s currently in my display name, sphexish, was coined by AI researcher Douglas Hofstadter to mean “behaviors that appear conscious but are in fact preprogrammed”

It is taken from the Sphex wasp which exhibits some complex behaviors - which are completely programmed into its genes

(Easy because the agent is kind of self-talking all along as it performs multiple steps, and so literally putting the message in the stream means it should naturally pick it up between steps.)

Intuitively I feel like dogs are conscious, but not as self-aware as humans. I don’t think ants are conscious, at least not on an individual level. My cells aren’t conscious, but I am.

That points me to believe that consciousness is systematic, that it arises from unconscious pieces.

A question: are animals conscious? Which ones? Are there animals that are conscious but less conscious than us? If so, what does that mean? How can we tell?

I asked my wife (someone lay but well informed) the difference between Machine Learning and an LLM. She defined Machine Learning as the learning process, and an LLM is an example of the outcome of the learning process. This is both accurate and not at all the distinction proposed here!