stcg

I don't know, but there is a small market for something similar: taking older cars (often oldtimers) and building in a new electric motor. My guess is that those cars will have little privacy problems.

I would love new open source cars by the way, that would be really cool

Exactly, it's just unfamiliar to most people.

I find Clojure (a Lisp) more readable than many other languages. It's the style I like for pseudocode. Because I used it a lot. It's familiar.

"Easy" is relative and doesn't mean anything without a subject. Easy for whom?

> proxy SDK monetization

Wait what? I never heard of that. I call that a botnet

Or make a different poll :)

https://www.youtube.com/watch?v=ahgjEjJkZks

Fun read! I love the depth in seemingly superficial texts. Sometimes I similarly ponder claims or questions in advertisements.

> Oxiis is a universal, friction-drive motor system that transforms any conventional bicycle into a smart e-bike. Its name fuses the Greek Oxis (agility) and Axis (pivot)—perfectly reflecting how its high-performance drive core delivers ultimate acceleration and instant, nimble responsiveness right to your ride.

This sounds like LLM-generated to me, but I find it difficult to point out why.

I have a similar experience with GPT 5.6 sol.

Nowadays I often prompt like "I heard there is also this different direction, what do you think about that?"

Another thing I do is asking the agent to make a decision matrix for choices. It's useful to discuss, give feedback on, and signals that it's a discussion, not a request for a particular direction.

It's then also easy to say: create a prototype for multiple directions so I can compare the solutions.

That way I choose the problem, I choose the solution, but the agent can help me discover solutions, make tradeoffs visible, and implement solutions.

That is also what I do now (both with NixOs and Guix Os), both for personal computers and for servers.

- I inspect the agent's changes, and only apply them - at once - if they are OK. So I have no half applied bad changes to my system, and I can catch critical mistakes before they are applied.

- I can roll back the changes by just doing a `git revert` and reapplying

- The agent cannot read secrets or unrelated data, just config.

- The agent gets the full configuration of all systems at once, without having to maintain parallel documentation (which can get out of sync) or rediscover each time from scratch (access my running systems, for example with `ssh root@server`).

- It's harder for the agent to miss some aspect of the configuration, because it's all in my dotfiles. If it's not there, it's nowhere

Wiki page for people like me who never heard of this test: https://en.wikipedia.org/wiki/Will_Smith_Eating_Spaghetti_te...

Yet another case of mistaking price for value.

When evaluating whether public money is well spent on education it must be more important how valuable it is to the public, not what the price for the work is to the individuals.

I like the "what if these workers stopped today" test:

Pick a profession. For example pick from 'trader', 'dentist', 'cleaner', 'sales person' or 'nurse'. Then imagine that all people in that profession stop working today.

How bad would it be for society? Is it better or worse than some other profession? Compare this to how well-payed the profession is.

I think this is a much better test for value to society than looking at what people get payed.

For example, I think it would be much worse if all nurses stop working than if all bankers stop working. Yet bankers tend to get paid more.

I like the word "botspeak".

Another example of typical botspeak is "smoke test". Why not just say "test"? It feels like a way of downplaying the ability to detect problems.

What makes you think it is slop? The emdash?

A letter from me addressed to the agent, describing me and the goals of the project. Not AI-generated, no code or commands.

Nothing specific to the technology, because that gets out of sync fast, or might steer the agent in the wrong direction.

Smart models can figure these things out themselves.

If for example the agent repeatedly wrongly guesses the command to run tests, I prefer to change the code so the wrong guess will actually work the next time. Or I will find what in the project lead to the wrong assumption and fix that.

When reading the article I thought of this programmer I worked with who actively promoted copying code because he thought it easier to copy than to work out a good abstraction. It did work at first but it started to massively slow him down after a while. The introduction of LLMs gave his project leeway, just in time I think.

I see now that indeed what this programmer promoted (copy immediately and don't try to abstract) is not the same as what OP promotes (backtrack on bad abstractions).

This is like saying "A slow leak is cheaper than a burst pipe"

Yes, okay. But with both you will have a bad time cleaning up.

There is a third option: good abstractions.

I did see this pattern described in the blog in practice a lot (and fell victim to it myself) and I think that in general this comes down to inexperienced programmers. Object oriented programming makes it worse.

Teaching these programmers that they should not abstract is not the solution. It is blocking their growth.

Teach them how to make better interfaces instead.

> That sure looks like software to me!

> But, at the same time, the user experiences it as a graphic. An animated GIF, for example, contains a small amount of code-like data to say how long each frame should last for and when to stop running. Is a GIF software? Is the basic circle above software? How much code do you need before something becomes software?

To me, from an ontological perspective, it's software if you wrote it like software: starting from the text form in a text editor. But it's an image if you made it in a graphical editor like Inkscape, just like how you could make an image in GIMP. The format just happens to be SVG.

I think this might also make sense from a legal perspective, as copyright law seems more focussed on the creation (how it is made / who made it / what was the purpose), and less on how it is used.

In the referenced benchmark GLM-5.2 (max) got 25% of all questions correct. GPT-5.5 (xhigh) got 57% correct.

https://artificialanalysis.ai/evaluations/omniscience

I'd much rather have some answer that I can verify than no answer to verify.

I don't want a model that says "I don't know", because I will verify the answer anyway.

Anyone know why there is a high frequency signal on top of the long term trend in that graph?

https://www.google.com/intl/en/ipv6/statistics.html#tab=per-...

I wonder what is the difficulty in rewriting strncpy uses that makes it take six years? Was it widespread? Or was it more of a long going effort, where it was only changed if there were some changes in the same file? Or is there some other thing that makes it difficult?

Beter wording: your brain is made for detecting dangers and attention grabbers are taking advantage of that

Other advantages to generating content: (1) fewer copyright issues. (2) No creators to pay, just GPU bills scaling with the use of the platform. (3) a much smaller critical mass.

Given these advantages I expect the current "social media" to be replaced with a new one, rather than them pivoting. The next big thing after tiktok might be something that only has generated content, where a last final bit of "social" is taken out of "social media".

I don't see consistent and correct as separate. Consistent is correct, and correct is consistent.

This sounds like how I think.

But for me, it often results in situations where I think much harder and longer than others but fail to act.

I learned to sometimes act instead of thinking more, because by acting I gain information I could not have learned by thinking.

Perhaps this human insight can be applied to working with LLMs. Perhaps not :)

"Let it crash" is a sentence that gets attention. It makes a person want to know more about it, as it sounds controversial and different. "Let it heal" doesn't have that.

One of the biggest usability problems with Python dependencies is that the name you import might be different from the name that you use to install the package.

So if you find some script on the web that has an `import foo` at the top, you cannot just `pip install foo`. Instead, you'll have to do some research into which package was originally used. Maybe it's named `pyfoo` or `foolib`.

Compare that to for example Java, which does not have that problem, thanks to Reverse Domain Name Notation. That is a much better system.

Ecclesiastes 1:9

    What has been will be again,
    what has been done will be done again;
    there is nothing new under the sun.

"XOR" as encryption probably refers to one-time-pad encryption, which is a very secure encryption strategy if done well, albeit often impractical.

(Note: I'm not saying anything about this specific code, which I have not looked at)

https://en.m.wikipedia.org/wiki/One-time_pad

Is bluetooth not usable anymore?

I don't think becoming vegetarian helps. From the article:

> The study found evidence that food processing is a likely source of microplastic contamination, as highly processed protein products (like fish sticks, chicken nuggets, tofu, and plant-based burgers, among others) contained significantly more microplastics per gram than minimally processed products (items like packaged wild Alaska pollock, raw chicken breast, and others).

Note the tofu and plant-based burgers.