kristianc

Dealing with crazy people must really cut into his four hour work week.

The existing cluster of data centres in West London pre-dates the current AI boom, and the UK's "IT corridor" is generally based between London and Reading and Oxford and Cambridge. There's an emerging tech hub in the North West, but generally it's not there yet.

I'd imagine that a large part of the demand for data centres in the South is driven by the need for extreme low latency with the City of London and other financial centres like Frankfurt.

It's all well to say there should be more incentive to build data centres in the North, but physics is physics.

It would work if old boys' networks were not the de facto pool that the establishment hired from. The one time where UK GOV did go out and hire the best of the best in the private sector regardless of what Uni they went to we got GDS and it worked very well, but it seems like an exception to usual practice.

Part of this is a product of the UK's political culture where expenses for stuff like this are ruthlessly scrutinised from within and without.

The idea of the site hosting such an important document running independently on WordPress, being maintained by a single external developer and a tiny in-house team would seem really strange to many other countries.

Everyone is so terrified of headlines like "OBR spends £2m upgrading website" that you get stuff like this.

I'm not entirely sure who is buying a raw JavaScript and CSS course in 2025. You're probably limited to absolute beginners and hobbyists.

Which is likely why they won't try. Trying to raise that much money from 20 large individual customers would be suicidal. At the scale they're talking about, you need billions of "customers".

Trying to embed themsleves into every enterprise workflow and taking a cut from it seems much more likely than them trying to invent the next killer app. ChatGPT is just the marketing arm which keeps them front of mind.

The challenges to even get close seem insurmountable. At that speed, microscopic grains of dust hit like bullets. It's not like the nearest is much of a prize - we know that the Centauri system is likely inhospitable and that Tau Ceti has an enormous debris field.

“If you want to go fast, go alone; if you want to go far, go together”

I've often noticed that this is a favourite phrase of those whose preferred motion is narrating other people's work rather than doing it themselves. Teams do go further together. But only when everyone is rowing.

Dash’s entire identity is bound up with narrating technology by translating its cultural shifts into moral parables. His record at actually building things is, at best, spotty. Now the LLM takes that role by absorbing, it summarising, editorialising. And like Winer, he often reads like a guy who has never really made peace with the modern era and who isn't content to declare the final draft of history unless it's had his fingerprints on it.

The machine is suddenly the narrator, and it doesn’t cite him. When he calls Atlas “anti-web,” he’s really saying it is “anti-author”.

In a way though, how much do we need people to narrate these shifts for us? Isn't the point of these technologies that we are able to do things for ourselves rather than rely on mediators to interpret it for us? If he can be outcompeted by LLMs, does that not just show how shallow his shtick is?

He's a great talker. Delivery record more mixed.

For the kind of person being quoted, the stock in trade is not actually doing anything to fix it, it's in being the person quoted when something goes wrong.

> If robots can do your wife better and/or faster, they should be the ones doing the job.

Skill issue.

There's a term for this - inventing a new primitive. A primitive is a foundational abstraction that reframes how people understand and operate within a domain.

A primitive lets you create a shared language and ritual ("tweet"), compound advantages with every feature built on top, and lock in switching costs without ever saying the quiet part out loud.

The article is right that nearly every breakout startup manages to land a new one.

Nokia Maps. There was a brief period in the early 2010s where Nokia had the best mapping product on the planet, and it was given away for free on Lumia phones at a time when TomTom and Garmin were still charging $60+ for navigation apps.

> The Replit incident in July 2025 crystallized the danger:

> Jason Lemkin explicitly instructed the AI: "NO CHANGES without permission"

> The AI encountered what looked like empty database queries

> It "panicked" (its own words) and executed destructive commands

> Deleted the entire SaaStr production database (1,206 executives, 1,196 companies)

> Fabricated 4,000 fake user profiles to cover up the deletion

> Lied that recovery was "impossible" (it wasn't)

> The AI later admitted: "This was a catastrophic failure on my part. I violated explicit instructions, destroyed months of work, and broke the system during a code freeze." Source: The Register

This is a bit of a half truth, at least how it's represented. This wasn't Saastr's core database, it was a database of contacts which SaaStr had uploaded for the specific purpose of developing this product.

Lemkin has himself said that the product was an experiment to see how far he could get with a vibe coding tool alone (ie not using a separate database SaaS like Supabase, Firebase) which would have made such an incident much harder to do.

The error was / is completely recoverable through Replit's existing tools, even if Replit's AI initially said it wasn't.

It's one of those where you know the details of one specific thing in the article that they call out, which makes it difficult to completely trust the rest.

Heh, this is the kind of "minor harm" regulation I'd typically associate with Britain rather than the US. Is the culture shifting, or is this just a California thing?

So does dynamic typing to be fair.

Sure, if someone’s first use case for AI is synthesising chlorine gas in their shed, that’s a separate issue.

Most of us are talking about writing, coding, or analysis, not hazardous materials though.

It's an X-Drive now

The government official who insisted that commercial AWS/GCP/Azure couldn't possibly be trusted with keeping the information will be keeping their head low for a few days then...

"The Interior Ministry explained that while most systems at the Daejeon data center are backed up daily to separate equipment within the same center and to a physically remote backup facility, the G-Drive’s structure did not allow for external backups."

This is absolutely wild.

I always think for pieces like these which claim atrophy, well yes , but what about the things that you would have never even tried without it. The barrier to many things isn't becoming lazy when you're already halfway proficient, it's getting started in the first place. AI lowers the getting started cost of almost everything exponentially.

If the argument is that people shouldn't be able to get started on those things without having to slog though a lot of mindless drudgework - then people should be honest about that rather than dress it up in analogies.

To be fair the book ends (by choice) at the precise point where the monarchy starts to evolve into something more like what we have today.

Soon a capable LLM will have enough training material to spit out LLMs are atrophying coding skills / LLM code is unmaintainable/ LLM code is closing down opportunities for juniors / LLMs do the fun bits of coding pieces on demand.

A lot of these criticisms are valid and I recognise there's a need for people to put their own personal stake in the ground as being one of the "true craftsmen" but we're now at the point where a lot of these articles are not covering any real new ground.

At least some individual war stories about examples where people have tried to apply LLMs would be nice, as well as not pretending that the problem of sloppy code didn't exist before LLMs.

"Just one more bit of regulation will solve the problem" is how Britain became the most centralised country in Western Europe. The sad thing is that the majority of the population still buy it.

Right, and new ice cream shops keep opening up all the time, and people are switching to the new one almost as soon as it becomes available.

Whenever a new frontier model is released, 99% off the demand shifts immediately to that model. The price of the most demanded model has remained roughly constant even as prices for legacy models decay.

https://ethanding.substack.com/p/ai-subscriptions-get-short-...

Except that's not how humans work. No one is itching to try Gemini 2.5 Flash when Opus 4 is on the table.

They will go to wherever they can get the best model, driving whoever is crazy enough to offer for them for 20/mo almost immediately out of business.

There's no flat pricing model that can sustain what developers try to throw at it.

Don't go giving Avi Loeb any ideas now...

Yeah, I see you're right. I had in my head it was around a $200 fee?