Your prompts, especially if they contain your entire codebase, are _way_ more data-rich than an airline phone call or 1920's novel. So they _are_ going to train on them, no matter how many checkboxes you tick to stop them. Which is both a commercial risk and a massive new attack surface.

It's not just software developers; if law firms aren't controlling any public LLM prompts _very_ carefully, they can expect some meaty client confidentiality suits. In fact any organisation in a competitive environment should be worried: Acme Bolts: "Write me a presentation for Zoom Construction". Beta Bolts: "Is Acme Bolts pitching to Zoom Construction?"

Maybe this is part of why OpenAI and Anthropic are finding demand softer than they would like. And why open-weight models that organisations can run on exclusive hardware are thriving.

I've lived with my floaters for twenty years or more, but sometime on a recent holiday the biggest one moved slightly to obscure the central vision on one eye much more often. Now when a floater in the other eye drifts onto the central area, I just can't focus. I have to flick my eyes to the side hundreds of times a day, and it's annoying enough that I'm considering a vitrectomy.

Maybe the techbros are playing irony roulette, seeing who can get away with the most outrageous hypocrisy.

I hate LLMs because their pushers are hurrying us into the next financial crisis, which will impoverish another generation.

I hate LLMs because they have captured pretty much all R&D investment when we have far more important and urgent existential problems to solve. The world is literally burning in front of us, but no time for that because an even bigger model might make 0.03% less mistakes.

The fediverse is a step in the right direction, and Meta charging may create another wave of converts there. It has a lot of growth pains to endure yet, but the ability to painlessly spin up your own instance could be very attractive to young people looking for their own non-corporate spaces on the internet.

We may also see some renewal via large companies (Meta in particular) imploding, from mismanagement and disenchanted users. My experience marketing a new product is that online advertising is completely ineffective now the web is filled with slop, no matter how well targeted it is. We've recently pivoted to optimise for word-of-mouth with orders of magnitude better results. I think any adtech company without a solid alternative profit stream is in for a rough ride (and no, AI is not a solid profit stream for anyone but Nvidia).

He believes Oracle has already signed it's own death warrant, and that Meta is close behind. MS, Amazon and Google have massive revenue streams to sustain them, but looking at the numbers, each has to earn from AI the equivalent of their existing real revenue. I can't see that happening.

And he believes from multiple perspectives of the data that Nvidea are either massively overstating their GPU sales, or that there are warehouses full of unused GPUs. There just isn't the energy capacity to run them all, let alone data centres to put them in.

Take a step back. What users want is to be able to use the machine they bought the way they want. The outrage is because Bambu are doing a bait-and-switch: selling an autonomous 3D printer, but switching to a 3D printing service. Enshittification pure and simple.

Apparently Zed was working on a better diff viewer, but that seems to have been shelved.

I choose to see every post on my (small country) instance, so if there is an echo chamber, it's an instance-shaped one. Which I like - I see the range of views prevalent in my small country. "Elite" posters depend on posting good content and are rewarded only by other people boosting their posts.

I tend to use Mastodon which makes finding a post's popularity a click away, and so emphasises posting for interest instead of outrage. This may also be an artifact of living in a small country that expects more civilised discourse from it's citizens.

Having no algorithm definitely makes the Fediverse more "boring" - I had to persevere after moving from Twitter. But I soon realised this was due to the lack of outrage, and that that was what I wanted, and what I was seeing was far more "real". Big fan now, and it's made my social media consumption a lot healthier.

Think about Tesla's pivot to "AI robots". My guess is that they'll get to something that can very slowly pick up a dropped sock and put it in the washing basket. But that it will fall over occasionally on the stairs, wrecking your kid's photos and the vase standing at the bottom, and dinging the wall. It might do a passable job of picking up the shards of pottery, but gluing the picture frames together, plastering the wall and repainting it... well maybe in in Elon's chemical dreams.

But now it works fine! Remote work is noticeably snappier than via mounting the remote server as a drive, and remote git seems to work nicely. A very nice Christmas present - thanks, Zed!

Good job Zed!

The fatal flaw of the open internet is that bad players can exploit with impunity. It happened with email, it happened with websites, it happened with search, and now it's happening with code. Greedy people spoil good things.

But I also really care about the quality of our code, and so far my experiments with AI have been disappointing. The empirical results described in this article ring true to me.

AI definitely has some utility, just as the last "game changer" - blockchain - does. But both technologies have been massively oversold, and there will be many, many tears before bedtime.