Binturong - Wikipedia

So there are a lot of possible reasons but...
en.wikipedia.org/wiki/Binturong

I went to college to be a writer or musician

but ye gods & little fishies, claiming a creative work is ready for the world is like taking a cheese-grater to the soul.

I was so much happier flipping to the "start tiny & iterate!" software eng path.

Anyway.
Huge respect to creatives of all stripes.

I keep resolving to start regularly posting creative work publicly

as a way to push back against AI music/art (I'm angry hearing AI music in cafés, and it's growing)

but my music never meets my taste, nor does the app I made to make bsky videos from audio

And oh, right: creative work is SO HARD.

It's absolutely not role-locked though; and many sw engs do settle into a comfort zone; I suppose it's just more obviously a danger for them.

But I think the best people in any role are always sketching out better maps.

This resonates — is someone actively growing & inking mental maps?

There's a point for many sw engs where they stop customizing things in their OS & dev tools— not because they mastered it, because it's ever-changing and a huge time sink

We do have to choose what maps to grow/ evolve, which not

And this is not my usual driver, but I'm willing to accommodate the vagaries of fate

I mean, if the scurrilous project to displace sciurus vulgaris has already triumphed in your locale, squirrels will be "to spec" but it's callous to crow about it

It may be painful to acknowledge, but widespread No Kings protests still failed to trigger any notable public policy shift, even in Derry, Maine and its environs.

Local pols floated a few trial balloons but that was It.

I'm already in the discord, looks like, though I so rarely have discord up I haven't even really looked around;

self-nudge to do that

V fair point — people as a whole are very bad at guessing what others want & how they'll navigate a change in their actual lives — everything AI seems to trigger particularly bad guessing

Yeah — tbh it can't be fast enough for a lot of these directions to deflate, and the AI investment FOMO panic to get less credulous

I think we should be upset about the dangerously broken govt (lack of) regulation, snake oil products, social harms;

I didn't even mention ecological concerns above.

It's not all going to vanish when the bubble pops, though. Only the useless part will pop.

I've been building software for almost 3 decades. You have to apply the same guidance as always, but some parts are far, far faster with partly-AI tools if you do.

The other parts are not faster, but hey, critical mind still engaged: none of the "new" problems are really new or intractable.

I'm v hesitant to join these debates, but more than one thing can be going on at the same time.

The US govt keen to not regulate AI at all is a very unhealthy force on competition

and lots of AI products are bad ideas, with broken UX.

Also, AI is a huge multiplier for some people & applications.

Worth noting that if we're talking about "regular people" then a tech like passkeys is *extremely* difficult to roll out effectively

Most tech is a nightmare for my neighbors already

& who has faith that anyone resists the "let's make it better for people with only our devices" temptation?

Money plays a huge role in waterfall, too — people with money saying "we want to pay a specific amount for a specific thing" and that sounds kinda reasonable

Then they give money to the people who say it can be done, not the ones who say it can't

(and then learn a painful lesson over & over)

(This all looks mushy & doesn't address the main issues clearly; I'm going to keep iterating until the line are cleaner; & keep evolving my own understanding. Need to keep researching the whole ecological side as well; it's hard to get anything like a clear view)

Worth saying also that the world is cruel & capricious, but all the good things we have come from people seeing they can help others and figuring out how to do that.

Dear reader, we see you are getting the large print copy, and that choice may be reminding you of your failing vision, diminished capacities, and inevitable mortality.

We know this may already be casting a pall over your morning

but reader, dear reader, we need to talk about this book — THIS book,

The website made me sit

and stare

This user totally ignored the "PDT" after the time

"Potentially Delayed, Thanks" is a common way to indicate that—

Dunno, no clue how to convince anyone to pay closer attention.

But the EU seems to be taking regulation more seriously; obv US federal is a clownshow but some US states are rolling out data center limits (a good step, imo).

I'm not going to drop into people anti-AI threads to argue it, obv.

A lot of the capabilities required also work decently well with models that run on a laptop.

What happens if the first bubble pops, but an actually-very-effective core remains & continues growing, and all the thoughtful, socially-aware people are still totally disconnected from effective regulation

There's a special kind of whiplash to signing off of work, where a handful of engineers have implemented a startlingly complete and test-covered, scalable product from hackathon to live customers in a few weeks

to social media, where the people I agree with about dangers also think AI has no value

but everyone who's just waiting for the moment it becomes obvious that no, the value simply isn't there, stays totally disengaged.

There IS a lot of silly stuff that's going to collapse hard; there should be some partial AI bubble pop

but nowhere near all of it

— & I also worry about impact on lots of human domains (creative, education, healthcare, legal/ criminal justice, etc)

and I do not trust the ethics, assumptions of how a healthy society functions, etc. trained into models, with good intentions or not

so I want to see discussion of regulation

All the "this technology isn't valuable" I see seem based on the first idea. Hesitating to dive into it, but really, assume *some* significant value exists, and it's greater than the heat of flaming hedgehogs running my underfloor heating.

My spouse is a novelist in the Anthropic class action suit—

Have been thinking about this for a while this evening.

Using an LLM directly — like, I ask a question, the LLM drafts a response that is shaped like a good answer, but may have errors — is weird & limited.

This isn't how any of the successful applications work though (not "popular", "successful")

posted the images in the wrong order, possibly

okay, so not even "on balance, this is more harmful than good"

but "zero utility can be achieved with this tech, full stop"