The most obvious reason for this seems to me to be that the data they were trained on likely overrepresented cases where the current choices were treated as unchangeable.

Actual case studies where they talk about what approaches were abandoned and why certainly exist, but I'd guess are much less common than posts presenting a problem as "these are my constraints, how do I meet them?".

It's not only Facebook doing it, but they are probably the biggest offender.

Edit: Okay, I was actually thinking of the Apple IIgs, which came out 2 years later. Both a similar design, though, and much nicer looking than the older Apples.

In the really long run that's of course true, but in the nearer term we are not going to have a replacement for jet fuel any time soon. So, for example, I personally would not speculate financially on this expectation quite yet.

"We run Gemma4-26B-A4B, distributed over 15 inexpensive off-brand Android TV boxes (~$55 each) that we bought and networked together, at 0.39 tokens per second.

[...]

A model that’s too large to fit on one device can be split into pieces, or shards, that run on separate machines that communicate with each other to collectively run the model. Machines that could run a significant model shard for a <30B model include smart fridges, smart ovens, automatic vacuum cleaners, and lawnmowers."

It is pretty interesting stuff. Still can't compete with a datacenter, but there's plenty of room for shenanigans.

I say the more the merrier, but artists who use AI in their workflows are going to be pressed about it - with negative connotations - for a while. Transparency is the best solution in my opinion.

My understanding is that 'emotional' sense emerged organically in LLMs and was not intentionally baked in.

As silly as it may seem, the ways emotions come through in our words do seem to have an effect on agents. Likely because they were trained on human writing, most of which cannot be separated from the emotions of the writers any more than your emotions can be separated from your logic & reasoning abilities.

This immediately invoked the old phrase, "bicycle for the mind" for me.

I've had success with agents finding old bits of media I couldn't find on my own through search engines, but I've never taken it to the project management level like this. Hard problems require more intensive scaffolding, I suppose! To me, the AI is more of a side story here considering everything OP did, but it clearly was useful, even essential.

I took it for granted, expected it to always be there-- wish I had something similar to point modern juniors to.

The obvious question here seems to me to be, what if they just burned their own biofuel? How much would that increase the cost/price?

At least some of these operations have to be running at some kind of net gain. It's really about price in the end. I've heard breaking down switchgrass in giant vats with microbes is going pretty well.

Eventually at some future point, we will need large scale biofuels at least for jet fuel. Because however long we might draw it out, we will run out of oil (better yet, we choose to stop using it, but, you know). The market will pay whatever it takes. The problem right now is that this seems so far off in the future and so impossible to 'time' that markets don't care yet.

I spent a lot of time cutting my teeth in MS Access, and would've done a whole lot more in VB6 if it didn't become deprecated so quickly. It's like a whole smooth ecosystem I expected to persist and improve just evaporated, all in favor of much more fiddly stuff or much more expensive stuff (or if you're really unlucky, both).

I would like to see a comeback in this area, with modern security features and reduced bottlenecks for scaling.

Except for those in power, of course.

1) The first mission would almost certainly be a set of robotic probes (especially in light of recent developments)

2) As controversial as it may be, we could genetically engineer descendants who are better suited for space travel

3) Cybernetics hold the potential to preserve the biological human 'essence' (brain and nervous system) while minimizing the amount of living material requiring maintenance

4) Hibernation is still a possibility, or at least we haven't ruled it out yet

When you're looking at long enough time spans - which we should have available if we manage to avoid extincting ourselves - many more possibilities emerge beyond simply "put a bunch of regular humans as we currently are into a tin can".