I suppose that's what happens if demand goes up for external reasons while supply doesn't follow.

I think we are at the beginning of a new golden age of software engineering, but not in the way we think. The cost of creating software might go down to zero with AI. That doesn't mean that no skills are needed. Quite the opposite. If AI would create outstanding code, then it would win every Kaggle coding competition on the planet. But it doesn't.

]1] https://github.com/storytold/artcraft [2] https://huggingface.co/blog/a2aprotocol/universal-modder-a-p... [3] https://jdsemrau.substack.com/p/jamming-with-jev-and-claude-... [4]

> The cost of creating software might go down to zero with AI

Keep hearing this, but I think people focus too much on only the first part of software engineering: the producing part. But one also needs to make sure that produced software works as intended… and here engineers (as of October 2026) are as much as needed as pre AI times, if not more (one could argue that more skilled engineers are needed to review the work of decent LLMs spitting decent software but with subtle bugs)

Great work, op. I have also had good experiences with Qwen3-TTS. Might be worthwhile to explore for everyone who wants to do longform locally.

Thanks! Currently, I'm using Gemini TTS and Kokoro TTS as a free option. Kokoro is fast and I can run it on my own server, but I'm not using it as a standard yet because Gemini supports more languages and has more voices available. But text-to-speech processing is the biggest bottleneck at the moment.

An LLM is just a brain in a vat. Our brains would also not reason in such a condition. Given the right framework that can do miraculous things. I am not saying that these systems are conscious, but they are able to abstract their context in a way allowing them to understand their own limitations.

https://www.finextra.com/blogposting/31255/adaptability-as-e...

Humans are also ethnically homogenic animals.

Makes much more sense described in this way.

Mistral's annual revenue (700M) is what Anthropic generates in 3 days. It will be extremely difficult to build new models with these numbers.

Mistral just needs to good enough category think all those flash models or Qwen3.8 27b which they sadly aren't at the moment, that plus being European lab will mean that they will have very nice business. Even now these SOTA models feel too overkill for most tasks.

The question is not, will mistral be able to create the best models (seems bloody unlikely). The question is, will mistral have enough expertise to dominate its business use-case, which is the model infra/deployment market within the EU (which is more plausible, with even a 3rd, 4th, 5th best model development team).

And GLM local models can eat Anthropic's lunch just like both eat OpenAI's.

Mistral's revenue is mostly B2B, and that's much more difficult to move.

Local models or independent hosting companies. Like Baseten, but there could be dozens of them.

if you have a computer to run said local GLM models one

The big US labs are naively hoping that compute power will always be their moat. I strongly suspect they are dead wrong, and newer architectures will require vastly less training data and outperform the brute force approach of US labs.

Arianspace launches from French Guiana while a French commune is not really close to the actual continent of Europe.

French Guiana is in the EU though. While Norway is not.

It is indeed a huge success but it also underlines how far ahead the rest of the world is.

I'd honestly rather like to use this opportunity to talk about that this is a success than returning to our regularly scheduled doomer loop.

A rocket launched into orbit from Norway. That's insane progress and worth celebrating.

> That's insane progress and worth celebrating.

A non-reusable launcher that can put small payloads into a Polar orbits only, due to launch site latitude and lack of an orbital bus. It's fairly 1960s stuff.

Did Europe have those capabilities in the 1960s?

Or is this insane progress for Europe?

only four countries in the world have private/commercial space launch capacity, and SpaceX towers over everyone else so I'd say this is doing pretty alright compared to the rest of the world

As EV’s have shown, I don’t think they’re gonna be ahead for long and long-term the current technology is just a steppingstone because damaging the ozone layer can’t go on forever with the current technique being used something else. Better has to come forward.

So what should they do nothing? They have to get started somehow if Japan and China can do it so can Europe, and I think long-term once again, Canada, Australia, and New Zealand won’t be far behind…

I think New Zealand is ahead. RocketLab was the second company in the world after SpaceX to get a privately developed launch vehicle to orbit, back in 2017. They've now had 86 successful launches from their NZ launch site.

That's essentially what they have done for military/defense.

The United States saves the day. France pretends they are in charge with grand speeches.

I was close to a European immigrant and 2 things she was shocked about:

How rich the United States was. Even our poor had more than her upper middle class family had.

How people in the United States do not take 2 hour coffee and cigarette breaks.

Is this the joke section of the comments?

The Taco section, with the way the present American administration has behaved. Most of our friends/allies are looking to protect themselves andbe prepared to go on without us. All it took was two years of buffoonery.

Utter bullshit.