Alan Buxton

Ok I give up. I'll read the New Yorker article about if Claude is conscious, given that you lot are all quoting it.

I loved Dungeon Crawler Carl and was v tempted by Alchemised, if only for the gorgeous cover art. Would love to hear how good it is

After hours struggling with getting Microsoft Copilot to help convert some Loop pages to Loop components I gave up and asked ChatGPT. It helped me straight away.

Go figure.

TBH this isn't a "ChatGPT good, Copilot bad" post.

It's "if one AI tool isn't doing it for you, try another one" post.

foreign [Music] Klangphonics recently turned up on my YouTube recommendations. Great musicianship, great sounds. Like the best bits of 1990's Trance. Love them. Here are some screenshots from one of their YouTube recordings with YouTube's auto-subtitles. At first, I found it entertaining how YouTube struggled to subtitle this music. After all, this sort of music isn't really about the lyrics. I've no idea why YouTube says "foreign".

foreign [Music]

Klangphonics recently turned up on my YouTube recommendations. Great musicianship, great sounds. Like the best bits of 1990's Trance. Love them. Here are some screenshots from one of their YouTube recordings with YouTube's auto-subtitles. At first, I found it entertaining how…

Nice subtle touch on this week's Private Eye cover.

Wow! Syracuse beats Perplexity! My side project made almost a clean sweep of the AI analysis of the latest head-to-head between Syracuse, LinkUp and Perplexity. Grok, ChatGPT, Claude and Gemini all favoured Syracuse. CoPilot preferred LinkUp. Feeling quite proud because middle of last year there wasn't much difference between one of these tools and another. To be fair, it was a very specific use case.

Wow! Syracuse beats Perplexity!

My side project made almost a clean sweep of the AI analysis of the latest head-to-head between Syracuse, LinkUp and Perplexity. Grok, ChatGPT, Claude and Gemini all favoured Syracuse. CoPilot preferred LinkUp. Feeling quite proud because middle of last year there…

Watching my kids get obsessed by the likes of Timberborn or Alchemy Factory or those Roblox Tycoon games and wondering about the appeal.

Then I remember how much Civ and Civ II took over my life in the 90's.

Everything that happens has already happened and all that.

nuclear take

a new quality of "good code" is

components are straightforward to modify or enhance without side effects via LLM

The AIs love my side project Or are they just telling me what they think I want to hear? I compared my side project, with Perplexity and LinkUp. Results are in this repo: news-comparison. Since the previous test run I've made a few updates, notably: Implement Typesense for all search Re-created Industry Clusters now there is much more data in When looking for what happened in a particular region, only look for events in that region, not events that happened to companies with a presence in that region…

The AI's loved my side project.

But were they just telling me what I wanted to hear?

I asked someone else to do the same comparison. His AI's gave him some different results.

Small Large Language Models, Claude and Typesense Some notes from putting together this little demo around Typesense. Claude did most of the heavy lifting, I only had to fix the following: it used typesense 26, not 29 Claude used a GET which didn't allow for a long enough query so I had to change the endpoint to one that used POST It picked a basic "getting started" language model rather than a multilingual one.

Small Large Language Models, Claude and Typesense

Some notes from putting together this little demo around Typesense. Claude did most of the heavy lifting, I only had to fix the following: it used typesense 26, not 29 Claude used a GET which didn't allow for a long enough query so I had to change…

About a million years ago I made 2 PR's to an open source library. One was to update an error in the docs. One was to fix a bug. Guess which one languished around forever and which one got dealt with?

Docs won't get updated unless the updates are happening in the flow.

The Next Nik? Lot's of talk about Nik Storonsky of Revolut leaving the UK for tax-friendlier climes. But what about people like: Sophie Tharnton Priya Nakamura Gabriel Hortley Dmitri Castellanos What about them? What about the people in the UK working on something today that will turn into as much of a success as Revolut has done in 10 years time? We can choose a replay of what we saw with Nik prioritising his tax planning.

The Next Nik?

Lot's of talk about Nik Storonsky of Revolut leaving the UK for tax-friendlier climes. But what about people like: Sophie Tharnton Priya Nakamura Gabriel Hortley Dmitri Castellanos What about them? What about the people in the UK working on something today that will turn into as much…

In Dungeon Crawler Carl, which is the definitive text on counting underground floors, when you go down the stairs from ground level you are on the 1st level (below ground).

By analogy, the 1st floor above ground level must be the first floor up the stairs.

I rest my case.

thanks to claude code i might actually ship the thing i wanted to make back in 2021 but didn't have the patience for. incredible time for people with vision and skill but low execution velocity

the best productivity system ever:

1. open a blank note/page
2. list all tasks for today
3. mark 3-5 most important with a star
4. complete those tasks
5. repeat steps 3 and 4 until you're tired

stop wasting time overthinking your "perfect" system

You read this stuff and wonder how people can be so dewy-eyed about "the good old days"

"Real Brexit hasn't been tried yet"

Might not be consistent but it is sane.

AI by the kilo One of my least favourite jobs is answering this RFP question: "How are you using AI in your product?" I understand the intent behind the question. As a buyer of software, you want some confidence that your vendor is up to speed on the latest technology. You want some confidence that what you buy now will improve with the times, or at least won't end up obsolete.

AI by the kilo

One of my least favourite jobs is answering this RFP question: "How are you using AI in your product?" I understand the intent behind the question. As a buyer of software, you want some confidence that your vendor is up to speed on the latest technology. You want some confidence…

I did jury service 3 times Lucky me. Jury service is high up the UK agenda thanks to I'm just a normal punter who has been on a jury three times, so I definitely don't have the big picture, but I have a view and this is my blog so I'm going to share it 😀 Things I thought were good: On each occasion there was one person on the jury who had some useful real-world experience to bring to the conversation and which affected the decision.

I did jury service 3 times

Lucky me. Jury service is high up the UK agenda thanks to I'm just a normal punter who has been on a jury three times, so I definitely don't have the big picture, but I have a view and this is my blog so I'm going to share it 😀 Things I thought were good: On each…

It is fair to say that everyone on the jury brings a different perspective.

One thing I don't understand is why there is such a veil of secrecy over what happens in the jury room. It would surely be helpful to be a bit more transparent about that.

For what it's worth, I've done jury service 3 times and I think it's a lot of pot luck

YouTube video by pikachuonetwo Alas Smith & Jones - 1986 - Episode 4 - PART TWO

Reminds me of the punchline to this old Alas Smith and Jones sketch: youtu.be/Xk-9oSAFyl8?...

What's frustrating about LLM text is it almost short circuits your brain. It's optimized to convince a tired rater and so it is maximally convincing when scanning

AI Slop and Hiring The Economist had an article about how, in the olden days, an employer could infer some useful things about a candidate from how they wrote their cover letter. In the age of LLMs, candidates aren't writing cover letters any more - the machine is doing it (paywalled). The conclusion: Employers, having lost a way of telling strong from weak candidates, have cut all new hires’ pay, and more often end up recruiting worse candidates. 

AI Slop and Hiring

The Economist had an article about how, in the olden days, an employer could infer some useful things about a candidate from how they wrote their cover letter. In the age of LLMs, candidates aren't writing cover letters any more - the machine is doing it (paywalled). The…

Running taught me pacing. Startups taught me pacing with other people’s ankles tied to yours.

"'What's going to happen' asks Professor David Zeltzer, 'when we have (computer) animation of photographic quality?' The prospect of the 43rd, computer-generated, series of Terry and June being made long after the two stars have retired is a prospect that could make us rue the day computers were ever allowed into television But also imagine all of the repercussions if a false computer generated image was broadcast showing Neil Kinnock being punched by Margaret Thatcher, the thought may be amusing but the consequences more than a little troublesome. You'll never know what is genuine television unless you're told. Could you ever believe news again? "

From 1988.

https://archive.org/details/crash-magazine-57?view=theater#page/n67

Try my new file format, OCSV: Oxford Comma Separated Values

Something I didn’t see regarding language models back in 2022 In Feb 2022 I was experimenting with GPT-3. I was curious at the time about what new avenues this might open up for beyond the fine-tuned BERT and RoBERTa models that I was familiar with for classification tasks. I was curious if GPT-3 represented any deeper knowledge that could help us from a product point of view, for example in finding new suppliers for various goods and services.

Something I didn’t see regarding language models back in 2022

In Feb 2022 I was experimenting with GPT-3. I was curious at the time about what new avenues this might open up for beyond the fine-tuned BERT and RoBERTa models that I was familiar with for classification tasks. I was curious if GPT-3…

Is that metaphorically speaking?