Chris Ashworth

"the primary purpose of the workspace status line is for posting on bluesky"

Lumping all of AI into one politically untouchable bucket is an unrealistic and sort of achingly bad move. There is no conceivable future technological infrastructure that does not involve high volume statistics.

But e.g.: are we building a future soiled by cars or are we building mass transit?

Tahoe is no doubt a UI disaster. But I'd still take a native Tahoe app over any Electron app that I have ever encountered. Electron apps are terrible.

My company has been building all native Mac apps for 20 years 🤷‍♂️

now that claude code exists no one is allowed to make any new Electron apps ever again

even if the code is sh-*t it'll still be better than an Electron app

Stars Hollow coded

ha, yes, this is me!

if any local news outlet is looking for a story ( @thebaltimorebanner.com ?) the therapist pipeline for the entire state of Maryland is currently backlogged for MONTHS behind a single woman who processes licenses and has what I can only describe as a very loose relationship to "doing her job"

thanks! i do some photography in the time between doing other things. for example, I take most of the show photos for voxel.org/events/

p.s. we're holding auditions this month

i took a break from work yesterday to get my head out of computers and AI.

I'm co-directing a production of Eurydice early next year, and I wanted to explore photos for that show. My co-director Tessara stepped in as a temporary model, but now I want to actually use some of these...

Just started to understand this today and it’s crazy!!

I think this captures something fundamental about what is happening in the world of "organizing electrons to make them useful"

people are just walking around in the world, not understanding that they can make software now.

like, the general public just doesn't get it yet.

Almost a full month ago I started trying to get @voxel.org set up with an Apple Developer account so we can sign and distribute software.

I have been through at last half a dozen support reps and still can not complete signing up, if you're wondering how the Apple Developer experience is.

I wrote a tiny little PHP script to show me when the next bus would arrive at the corner by the apartment, but with my ADHD I would often miss it anyway. For years afterward the sound of a bus leaving stressed me out real bad.

I was clearing out old files yesterday and found the rental agreement for the apartment I was living in when I wrote the first QLab code. They've fixed it up since we lived there (and now it costs a lot more! Listed for $2000 a month now — we paid $695). www.apartments.com/115-pine-st-...

skipping over the part about how the *brain* of the robot running in the Linux computer that is running in my Apple computer is actually a cloud of untold numbers of *other* computers that my local robot is talking to over the phone. (to keep things simple)

trying to explain to my younger self who is using an Apple II that his older self will be writing software for an Apple computer by talking to a robot that is running in a Linux computer that is running in the Apple computer

Anecdotally (via myself, some friends, and some posts I've seen online) a lot of developers right now suddenly have:

A) an incredible new level of output
B) intense headaches

the new threshold for "should i write a custom program to solve this problem?" is making me seasick

in the Before Times it was "almost certainly not, but perhaps once in a great while, after pondering it and preparing for the commitment"

now, for any given program, it's "lol probably???? try it!"

The heading of a paper entitled "Empirical evidence of Large Language Model’s influence on human spoken communication"

Noting with amusement:

In Star Trek, being part of The Borg means that "you" think like "it".

(In seriousness, though, this doesn't seem novel; humans have been thinking like the stuff we expose ourselves to forever.)

arxiv.org/pdf/2409.01754

I used to tell people that you could make an impact by becoming an expert at two things. It didn't matter *which* two things, as long as one of them was programming.

LLMs at least soften the barrier to expertise. All changes are both good and bad; let's find & support the ways this one is good.

Of course, providing all the inputs you care about is famously difficult.

I am not personally a fan of the "don't even look at the code" camp of LLM generation but I think it's intellectually honest to note that if you provide, substantially, every input you care about to a program, and it produces correct output for all of them, then it is, by definition, working.

Dario Amodei — The Adolescence of Technology Confronting and Overcoming the Risks of Powerful AI

I can see why this guy — the CEO of Anthropic — was either kicked-out-of or voluntarily escaped OpenAI.

As my teammate phrased it: having both vision and ethics can be a huge buzzkill.

www.darioamodei.com/essay/the-ad...

the closest reference points I have to "how this feels" is "early pandemic" when we were frantically trying to learn How To Survive. or maybe "an exciting high seas adventure, but also constantly seasick"

grateful to be seeing this in the discussion; was wondering if it was just me

I'm curious if you've dug in on using the recent coding models? There's certainly a wild amount of hype floating around but after exploring it earnestly last month I feel deeply unsettled to report that I think the quoted OP here is misunderstanding a real thing that is happening.

Terry Pratchett is probably my favorite author so if this little experiment is even remotely like something he would find appealing? what a wonderful thought that is to me (and how lovely is this quote, which I had not yet seen)