whazor

This will make it easy to add Pi.dev to a web application and connect it via WebMCP or a backend MCP.

Meanwhile Fable 5 and Opus 5.5 are in a different league.

Nixpkgs is quite explicit about how to produce the distribution. You can rebuild the entire distro from source and backup the source code for every package.

As a normal NixOS user, I know how to do this. It would be quite some effort, but not especially difficult. Additionally, for a government they have the servers to set up a build farm.

They previously used openSUSE and then Fedora. But both have companies behind them, where-as NixOS does not. And with NixOS they are more sovereign.

There are two problems that arise from cheap code:

1. You get complexity back at the speed that you add code. Also the complexity compounds.

2. As a human you are still responsible for your code, and you need to understand how the important parts of the code base work.

When you combine these problems, you get a complicated code base where you don't understand the important parts.

That is why software design is getting a comeback. It actually helps with both problems. Less complexity and more understanding.

Software engineering is so back.

When designing systems, you want the important details to be right. Especially with authentication and authorization.

From an architecture level, you can know which classes are important to review and which ones are not.

It is insane from a consumer point of view. Luna is cheap and smart enough to do many agentic tasks. Cheap enough so that you can put it on a website without auth.

The Jev model is economically, but also in terms of compute, a much more efficient model. A normal LLM goes token by token, each token in a separate step. Whereas Jev just returns all the results the first round. So it is much better at classification than LLMs.

Compared to traditional ML classification, Jev works without training, like a LLM.

MCPs are winning because within the ChatGPT and Claude apps, there are Plugin stores. These plugins are one-click installation MCP servers, with support for authentication. This is what business users are using.

The most interesting use case in my mind is skipping law suits. Obviously you need lawyers in court. But lawyers are people you are basically paying to fight for you.

Instead, if you resolve your dispute outside of court, you don’t need a lawyer. If both parties use ChatGPT to find the relevant laws or read contracts, they could come to an agreement without expensive legal fees.

Given the approach from the article, you can commit the hints to git and run tests for verification. The model would be used during coding.

A question I have, with the type { output: string }, would the model not become a LLM? And if it does, shouldn’t it cost the same as a LLM for output?

There are also people arguing that children should have no internet access because its harmful.

But children do have access to internet and often with some protections.

This is also how I feel about AI, where a full ban is the lazy way out. Similarly like the internet, it is not possible to 100% protect a child every potential harmful AI interaction.

This voice model can do tool calling. So as long you give it powerful tools, I believe it to be useful.

I can understand the concerns but at the same time it feels weird to deny young people the ability to make their ideas reality. I got proper internet access when I was 12 and that is how I taught myself how to program.

Once you know this, you cannot unsee this in every AI design everywhere

I'm confident that Muse can do lots of agentic tasks succesfully for normies on outdated models. Like buying sneakers, booking appointments, dealing with government forms.

Yeah and you probably have to. An ETF easily has >1000 different stocks and even being weighted. So it has a completely different risk appetite by being so averaged.

But ETFs do have to follow particular rules defined by their product description. So it is still interesting to benchmark against.

The 'edge' is holding the investments over long periods of times. Agents are merely automating the portfolio managing part for lower costs.

It would be more interesting to compare trading agents with index tracking ETFs. The better version of an ETF could maybe be a model where you zoom in on the companies and add/remove to your portfolio on the company related news, but keeping a broader portfolio.

Maybe agentic trading still performs worse than ETFs. But alternatively, if it were meaningfully better then it would be okay to opensource, similarly how ETFs are publishing their portfolios.

It makes no sense to train frontier models from scratch anymore. The best frontier models are only a half year ahead of Chinese open models. In this regard Anthropic and OpenAI are also in a bad spot when they waste so much compute on training models.

An important factor is that fine tuning existing open models is incredible cheap. You can easily change any cultural biases if you want a model to be 'sovereign'. And Mistral could combine that with their custom data sets for their enterprise customer needs. Mistral still trains their own models, but they also seem to offer fine tuning existing models.

With model weights being commoditized, another differentiator could be deploying efficient inference chips, especially if you combine it with a developer ecosystem for vendor lock-in. That is why it is interesting that both Samsung and ASML are investors, since they are companies that could make a difference in this area.

Is the article inventing any new words?

So many moving parts that can break. I am already sad my robot vacuum cleaner broke (it's the moving lidar that broke). I have to send the entire thing including dock back to the manufacturer (very expensive).

My preference would be ifixit level guides where you can easily replace parts yourself. Then it would be perfect if the same servo is used everywhere so I can have 3 or 4 stored somewhere.

Alternatively, AI teaches you lessons about technical debt, code maintainability, and architecture faster too. Traditionally it took at least one or two years to really bump into those problems. Now any student can get into these problems within one or two weeks.

The browser mode is great, except its fully banned by Cloudflare. I tried cancelling a phone subscription but Cloudflare stopped it.

My analysis is that Stripe wants to own the costs of running a startup. They would get a complete picture of both sides of the business: what money comes in and where it goes.

There are also a couple of advantages. They can take money directly from revenue before it leaves Stripes and without any processing costs. They can also invest into startups through credits and financing. And finally, their exposure to bankruptcy risk can drop as well.

I think it will happen.

And EU regulations are generally more two-sided where lots of innovations can happen by smaller marketplaces as well. Not just Amazon.

My robot vacuum cleaner broke. The lidar sensor is broken. The manufactures asked me to send it to their repair center but the whole thing, including the dock. This is a major pain to do so. I would much rather receive a new lidar sensor with some instructions on how to repair it myself.

Create new chat after step 2, then ask AI to fix the spelling / grammar of those three sentences

Fastmail search sucks

But now I use the MCP and that seems to resolve all my search problems. Now I let an agent search. Their MCP can still be improved though