Artificial Activities on Tales from the Bitface

Claim Chowdering Gruber's Claim Chowder

Sat, 14 Mar 2026 00:13:58 +0000

John Gruber makes a ridiculous assertion, or so it seems to me. In criticising Dario Amodei, the CEO of the AI startup Anthropic’s claim that ‘AI, and not software developers, could be writing all of the code in [their] software in a year’, Gruber takes things the other way:

It may well be true that 90 percent of the lines of programming code that are written today, Friday 13 March 2026, will have been generated by AI. If anything, it’s probably a higher percentage.

This seems like nonsense to me. Certainly AI-generated code is being created, and some of it released. But I work in software development, in a real company making real software that moves people’s money around. There’s some experimentation going on, people will use it to try things out, or better understand things, as I mentioned a few weeks ago. But there are millions of lines of code out there being written and managed every day by real humans.

And when you’re working in a highly-regulated industry like the payment card one as I am, or medical systems, say, it seems unlikely to me that we will ever let significant applications into the world if they were not written by humans.

Maybe I’m being naive, at least by saying ‘ever’: if there’s one certainty it’s that things will change. But the idea that we’re already above the 90% AI-generated mark? Sure, Anthropic are likely to be at that level. They build these tools. Eating your own dogfood and all that. But for normal, day-to-day development? It just doesn’t ring true to me.

Plus, of course, software development is about a lot more than writing code. But that’s a discussion for another time.

Wed, 11 Mar 2026 21:24:13 +0000

In more ‘AI’ nonsense, Grammarly is giving bad advice and tagging writers' names to it, without paying the writers or even getting their permission.

I tried Grammarly a few years ago and hated it, but that was long before the LLM boom. This is beyond unethical.

Wed, 11 Mar 2026 21:01:00 +0000

I find it deeply weird and surprising to read of authors claiming as ‘mine’ images they requested, or copied and manipulated using ‘AI’. The kind of claim quoted in the linked piece, that ‘it’s all mine.’ When it plainly isn’t.

Writers, you’d think, ought to understand that words have meanings.

Thu, 26 Feb 2026 19:10:07 +0000

Ed Zitron’s latest, On NVIDIA and Analyslop, is very good on the current state of some financial stuff related to ‘AI’. It’s also good on how much more complex software development is than the ‘vibe coding’ believers would tell you:

Software is a tremendous pain in the ass. You write code, then you have to make sure the code actually runs, and that code needs to run in some cases on specific hardware, and that hardware needs to be set up right, and some things are written in different languages, and those languages sometimes use more memory or less memory and if you give them the wrong amounts or forget to close the door in your code on something everything breaks, sometimes costing you money or introducing security vulnerabilities.

Automatic Introspection

Sun, 22 Feb 2026 19:31:56 +0000

One of my concerns about LLM-based ‘AI’ systems and their ability to generate code has always been the idea of people creating software they don’t understand. It doesn’t matter too much for an individual’s personal projects, for little utilities, and so on. But in a professional context, such as my work, it borders on the terrifying.

It can be hard enough, even as an experienced programmer, to understand the intent of an unfamiliar codebase — hell, sometimes it’s hard to understand code you wrote yourself a few months or years ago. How much more will that be the case when the code was generated by an automated process that was prompted in turn by someone who didn’t really know what they were doing?

That concern remains, even as I read stories of people getting Claude to write code, and getting it to fix it when it doesn’t work at first. In the long term — maybe the medium term — even bug fixing might be done by prompting.

But from a couple of recent experiences, one of my concerns has moved to something perhaps more familiar to non-programmers. People crafting English text that the reader doesn’t really understand.

And maybe that they, the ostensible writer, don’t fully understand. We’ve lived through decades of business jargon filling our textual brain space, read endless corporate bulletins where we wondered whether the writers understood what they were trying to say. And now the power of automated obfuscation is available to everyone.

Ablative Irony

Wed, 18 Feb 2026 08:54:36 +0100

Also via Kottke comes this article by Claudio Nastruzzi at The Register, where he talks of ‘semantic ablation’ in text generated by ‘AI’:

When an author uses AI for “polishing” a draft, they are not seeing improvement; they are witnessing semantic ablation. The AI identifies high-entropy clusters – the precise points where unique insights and “blood” reside – and systematically replaces them with the most probable, generic token sequences. What began as a jagged, precise Romanesque structure of stone is eroded into a polished, Baroque plastic shell: it looks “clean” to the casual eye, but its structural integrity – its “ciccia” – has been ablated to favor a hollow, frictionless aesthetic.

It’s about how LLMs — probability-based machines, after all — tend to push text in a generic direction, away from a writer’s unique voice, towards a common mean.

So let’s all not do that.

The irony is, I was trying to look up an unfamiliar word in that quote — ‘ciccia’. The dictionaries installed on my Mac had nothing useful, and nor did Wikipedia. DuckDuckGo’s search only came up with uses of the word as a family or brand name. I used the ‘!g’ syntax to send the query to Google.

It must be the first time I’ve had to do that in quite a while. I’ve heard people mention — complain about — the ‘AI Overview’ the Big G provides, but I’m not sure I’ve ever seen it before now. But it was what had the answer:

informal Italian term for meat or, idiomatically, body fat (flab).

Clearly Nastruzzi is using it as we might say ‘the meat of an argument’, or similar.

Google’s AI thing does not cite its source, though, and none of the next few search results give a reference for that use in English, though one is to the meaning of the Italian word.

Anyway, my recommendation to all fellow writers, would-be writers, and people who want to or have to communicate by writing: express yourself. Don’t let American machines do it for you (and use as many em-dashes as you need, as I have done here).

Good Programming Test

Sat, 14 Feb 2026 18:09:26 +0100

There have been a couple of interesting pieces in the ‘AI’ space over the last few days.

In a long Twitter post Matt Shumer writes breathlessly about how AI is changing software development, and how it’s going to change everything else. It feels simultaneously celebratory and sort of panicky.

It’s had a lot of pushback, including this piece from Om Malik, which has the advantage of linking to several other pieces that comment on Shumer’s. Shumer doesn’t have the luxury — or rather, the standard, decades-old convention — of linking to the things he refers to, because he’s writing on Twitter. Which has added the capability of publishing long posts, at least for blue-tick-verified users, which is good; but has not seen fit to include standard web feature like links, which is bad.

For me, he shoots himself in the foot with this exhortation:

If you’ve always wanted to write a book but couldn’t find the time or struggled with the writing, you can work with AI to get it done.

Ahahaha. No. That’s not what writing a book is. Writing a book — the clue is in the term — requires writing, which takes time, and craft, and practice… Getting your friendly neighbourhood AI to do it for you is very much what I referred to as cheating.

Anyway, both are worth a read, and the others Om links to, too. For me, they have contributed to a surprising change I’ve been noticing in myself. You’ll recall I wrote about feeling revulsion towards the modern AI situation. I find that feeling is fading, replaced with a sceptical interest.

Of course, I can no more explain why that change is happening, than I could the revulsion in the first place. Perhaps it’s just the ongoing piling on of the conversation. If anything, I’m getting bored of hearing people talk about AI on tech podcasts.

However, there is a change I’d like to make to some code at work that I think might work well as a test for possible uses I might have for the technology. It’s not a change to how something functions, just to how that functionality is implemented. It’s not essential, but it will make some things easier in the future.

And since we’ve got this GitHub Copilot thing, and it has access to various models, I think I’m going to give it a try next week. I’ll report back here.

Asbestos Intrusions

Thu, 12 Feb 2026 19:04:15 +0000

AI is the asbestos in the walls of our technological society, stuffed there with wild abandon by a finance sector and tech monopolists run amok.

Cory Doctorow’s latest piece is the script of a talk he gave on ‘AI’, or more specifically, ‘how to be a good AI critic.’ He’s writing a book on the same subject.

I found it weirdly comforting in one specific area. That of the supposed copyright-infringement of the training of LLMs. Cory explains why it did not, in fact, infringe copyright:

First, you scrape a bunch of web-pages. This is unambiguously legal under present copyright law. You do not need a license to make a transient copy of a copyrighted work in order to analyze it, otherwise search engines would be illegal. Ban scraping and Google will be the last search engine we ever get, the Internet Archive will go out of business

And he goes on from there, explaining why the subsequent steps in training also do not infringe. Some would disagree, of course, and many would say they put their work on the web with a ‘Not for commercial use’ type of licence, such as a Creative Commons one.¹

Which is fair enough too. I don’t think many would disagree with the idea that using the web to train these things was unethical; even more so with using pirated books. But it wasn’t strictly in violation of copyright (at least the current state of US copyright).

Why do I find that comforting? What I mean is, it removes or slightly reduces one of the reasons to be opposed to, or appalled by, these prediction machines, which I alluded to in one of my earlier thoughts about the matter. And in doing so maybe helps me in my quest to understand my own feelings, by at least reducing the number of things I have to consider.

Something like that, anyway. Read the whole of Cory’s piece, it’s very good.

I have done so myself in the past, though my site doesn’t currently show any licence. ↩︎

Less Like Manufacturing

Sun, 08 Feb 2026 22:18:03 +0000

Mitchell Hashimoto writes, in My AI Adoption Journey, of his process of adopting ‘AI’ agents in software development.

He describes himself as having been ‘a heavy AI skeptic’. My question, then, is, why did he force himself to try it?

He perhaps answers that when he says:

Adopting a tool feels like work, and I do not want to put in the effort, but I usually do in an effort to be a well-rounded person of my craft.

OK, I guess. But at the end he writes:

I’m a software craftsman that just wants to build stuff for the love of the game.

This feels relevant to what I was saying yesterday about tools like this feeling like cheating. If you’re only doing it for fun, as a hobby, then nobody’s judging you but yourself. But the craftsmanship, the writing it yourself: isn’t that the point?

If you did woodwork as a hobby, say, you could doubtless make a chair faster with an automated factory of some kind. But wouldn’t you lose the fact of being a craftsman, then? The very reason you took it up in the first place?

Generalised Philosophy Talk

Sat, 07 Feb 2026 20:26:42 +0000

Much of what we see written or hear podcasted about LLMs is how useful they are for programming. Coding, as they say, which may or may not be a distinction without a difference. I’ll come back to that, perhaps.

At work, as I mentioned briefly in a footnote last time, there is some pressure — or not as strong as that, let’s say wishing — from some elements of management, for us to make more use of ‘AI’. It’s a payment-card company. A regulated industry. It seems vastly unlikely that there would be a place to connect our customers' data, for example, to a US-owned (or Chinese!) probability machine.

But in the process of our work, internally; in software development, for example…

I resisted — or just ignored — the suggestion that we activate GitHub Copilot in our code editor, for a long time. And then when I decided to do so, to try it, I found I didn’t have the permissions. So I left it. Eventually, though, the permissions got sorted out, and I gave it a try.

There was a piece of code I didn’t fully understand, so I asked the chatbot built into IntelliJ to explain it. And, with a bit of back and forth, it did. It was nothing I couldn’t have found out by searching, but it was quite effective.

So I activated the auto-completion feature in my editor and let it go for a while.

Reader, yesterday I deactivated it.

It had become so annoying, so intrusive, I couldn’t stand it any more. Whenever I tried to type anything it would pop up a suggestion. Often it suggested exactly what I had been going to type anyway. Which is not the worst. We’ve had autocomplete in editors for decades, probably, where the editor will suggest the name of a class, method, etc, and it saves a lot of mistyping.

This new thing, though; this was suggesting whole blocks of code, practically whole methods. And worse, it was suggesting text. You go to type a comment, and it proposes something very close to what you were going to write. Or something quite different, but either way, it is not what I want.

They say having Claude Code or something to help you is like having a keen junior developer at your side. Fine, but if a junior developer kept trying to type what I was typing, I’d tell them to sod off and get on with their own work. Stop trying to do mine for me.

All of which leads me to another of the things about ‘AI’ that has bothered me since I first heard of people using them to help with work in this kind of way.

Maybe this is a silly concern, but: it feels like cheating. Getting someone/thing else to do your work for you? That’s not right.

Aging Inquiries

Fri, 06 Feb 2026 20:23:17 +0000

When I first started hearing about the modern suite¹ of ‘AI’ systems, I had a reaction. Now, I’m a computer programmer, and a science-fiction fan, so you might imagine my reaction being one of interest, intrigue, fascination; a desire to explore these new tools, toys, entities, whatever they may be.

No. Not so much.

My reaction was something I’ve been trying to understand and explain to myself ever since. It was an emotional reaction, not a rational one. I’m not talking about concerns over environmental issues, power and water consumption, or even the copyright violations of training the things. Those came later. My initial reaction was a visceral rejection. The closest I can come to a one word expression of it:

Revulsion.

Why would I feel that way? Why such a strong reaction, why such disgust? I don’t know. I’ve been struggling with it all this time, which is why I haven’t written about the subject.

I mentioned this at work² one time, and one of my colleagues said maybe my reaction was to some extent provoked by SF’s history of AIs gone evil. It was an interesting thought, but didn’t feel quite right. After all, my favourite SF novels, Iain Banks’s Culture books, have machine intelligences that are thoroughly good. Or devious, cunning, scheming, and all sorts of other things humans can also be. Chat GPT is not Skaffen-Amtiskaw. But what is?

I sometimes wonder if it’s just age. Maybe if I were twenty years younger I’d be diving headfirst into the afore foot linked pileup. But I don’t know. Which is going to remain a common theme of these pieces.

Tranche? Cluster? Pile? What is the collective noun for LLM systems? ↩︎
As you might imagine, there’s a certain amount of pro-AI talk at my work, though from what I’ve seen, mainly from management. I’ll have more to say about that as I write more of these pieces, I should think. ↩︎

Little Lost Machine

Tue, 03 Feb 2026 19:43:23 +0000

A little while ago, which turns out to have been June 2024, I microposted saying I ought to write about my thoughts on the current state of what people like to call AI. LLM-based prediction machines, some might say. Then about a year later I briefly wrote again about my negative reaction to the whole idea.

But I didn’t go into detail. And I’m still not going to; at least not today. I have several thousand words of attempted essays, if that’s not a tautology¹, wherein I try to understand my own thoughts and feelings.

And time passes. And the development of the things is lightning fast. It’s a moving target that annoys me.

Still, I do have thoughts. And feelings. And the best way to understand them is to write about them. And the best way to write about them is publicly. Maybe. So I’m going to try writing about them here. A series of short posts around that theme. This is the first.

Maybe I’ll give them their own category, though I have too many categories as it is. I discovered it’s hard to search my blog for ‘AI’. Micro.blog’s search is good, but that’s just such a common set of letters. Weirdly, it brought up all my Crucial Tracks entries, as if it was also finding the ‘IA’ in ‘crucial’.

What with ‘essay’ originally meaning ‘attempt’. ↩︎

Wed, 30 Apr 2025 19:52:40 +0000

I still don’t understand why AI gives me such a visceral negative reaction.

The intellectual reasons for concern are well known.

But right now, I just wish apps would stop adding AI and trying to tell me it’s great. I’m looking at you Raycast, but you’re just the most recent culprit.

Thu, 06 Jun 2024 17:37:17 +0000

I keep thinking I should write about the current state of what we are calling AI. Trouble is, I still can’t quite decide what I think about it. Or why it makes me feel the way it does. Or even what, exactly, that way is.