I see so many places that could have roundabouts instead of lights. Plus, if the maximum speed inside my town was lowered to 40km/h, we could have shorter yellows. You can never reach the 60km/h limit for more than a few seconds anyway.

Speaking for Johannesburg, we can't have shorter yellows because too many people treat them is "last chance to speed up to make it through before it's red for too long!" and JMPD is beyond useless at law enforcement.

I used a machine that had 8MB of RAM and a huge b&w monitor. One day I found out it would crash if your process forked more than 127 times. These things were so cool at the time but are nothing compared to a Raspberry Pi.

It's absolutely crazy to realize that even a Pi 400 running from MicroSD would smoke classic workstations of yesteryear, and yet by today's standards it's slow.

I tried this with my kids, but I required them to also produce a one page essay on each book. They ignored the deal. One of them actually reads and likes reading, but he won't be bothered to write about what he has read (unless it's required for school).

There's a book called Kuru by an Australian doctor who worked in PNG and studied the disease. IIRC, it's by Michael Alpers. It's a fascinating book, and it's a mixture of adventure, medicine, and anthropology. I highly recommend it if you can find it.

Neither the West nor the USSR grew coffee, though.

The West wouldn't trade (or let developing countries under its influence do it) and I bet the USSR couldn't procure enough for itself and the DDR.

This week I helped a colleague with a bug. He used an LLM to sort it out before telling me about it.

I'm old fashioned, so I looked at the code, saw that it used an old lib and then I searched online for a newer version. Lo and behold, the bug had been ironed out.

AI makes it really easy to not think properly about your task.

That same story happens without LLM involvement. I’ve also seen LLM suggest updating a lib. So unfortunately nothing about this story speaks to a difference between LLM and human capabilities.

Exactly. I feel that agentic programming is largely turning what would have in the past been actual understanding into some kind of fluffy superficial pseudo "understanding" that's only barely good enough when assisted by AI.

how did his AI solve the bug? I've had cc recognize a bug was in an outdated lib and then update it to resolve the issue. Granted it made that determination by finding the source code to the lib, identifying the bug, and then verifying it was fixed by looking at the source of the updated lib.

AI told him to use an alternative function which was similar but didn't have the same bug.

AI providers are in fact incentivized to have you not think about the task. I’ve found that they automatically do more for you, and frankly, a lot of the time I don’t want it to do more for me. Ask me a question rather than burning a bunch of tokens to answer something I already know of the top of my head.

The agents also overload you with walls of text, always including lots of less relevant information. It's like they want you to give up and let them drive.

This sort of thing should be labelled.

Just as food has warnings (too much sugar, too much salt, etc), so electronics should be labelled: "requires online registration" and "captures conversations" and "scans your network".

That's not enough. We don't allow food companies to just slap on "contains lead" and keep selling poison. And if every device has this label it just becomes useless.

This is not a case of maths giving you the Blues.

Someone at work said something along the lines of "so AI is going to write the documents and then people will use AI to summarise them". I think this is a good opportunity to rationalise bureaucracy but I fear we will only mindlessly automate it.

With many bureaucratic systems AI seems to be used to build a better horse when cars and tractors already exist. Everything was built for a paper based world and was never designed with data or automation in mind.

I see it as the first step towards removing entire workflows like this.

Once the humans are removed, there is far less motivated reasoning to keep that process alive if it’s just robots burning tokens passing paper back and forth.

I will likely somehow be proven wrong, but it is one of the main uses I am optimistic for in the AI space.

Yes that is a good point and AI also gives us tools to make traditional systems more quickly. It could actually deliver on some of the original benefit promised by computing.

When I first found out about Tcl/Tk, it was used to give a GUI to programs written in C. I've used Tk with Tcl and many times with Perl too.