For the AI frontier, I find my windows PC just about useless unfortunately. Too much tooling and package doesn't adapt to WSL+windows host well. I've shifted my entire dev experience to my mbp which used to be my backup. Can't imagine the new generation of vibe coder will even consider a windows box.
What would it look like to regularly react to source data changes? Seems like a big missing piece. Event based? regular cadence? Curious what people choose. Great post though.
For code specifically this is the hardest part — the "source data" (the codebase) changes constantly with every commit, but the AI config files that
describe it don't update automatically.The approach that works best is AST-diffing rather than hash-based reindexing — you can detect semantic changes (function renamed, interface deleted) rather than just textual changes, which gives you much more precise invalidation signals.
Depends on the use case, ie frequency and impact of changes.
Typically you would have a reindex process, and you keep track of hashes of chunks to check if you’ve already calculated this exact block before to avoid extra costs. And then run such a reindex process pretty frequently as it’s cheap / costs nothing when there are no changes.
The people for whom I've seen "coding is the hard part" are typically promoted out of the way or fired. They never entered a flow like those who considered it easy and addictive. The latter are the pillars of the eng team.
My feed isn't as bad as this one, mostly current events, tech, music, politics which are my interests. Trolls/ai/bots are everywhere, but so are people callling it out, so if anything I would guess engagement is up. To be fair, my politics seems to be around 60/40 agree/disagree with my political preference which I actually think is a massive improvement over what it used to be which was 90% agreeable to me. I enjoy engaging on pages of the opposing view.
Why are we still supporting this person? His cars are being outclassed internationally and he's directly meddling in this countries politics. He spectacularly failed (or wasn't it blatantly misled) the CA government with regard to the tunneling, and damaged the public sector while shutting down oversight and regulatory bodies against his companies.
Where is the benefit? These awesome tech demos? It just screams charlatan to me on an epic scale. I see no reason a government shouldn't step in to assume control if its "too big to fail".
Is text that perfectly with 100% flawless consistency emulates actual agency in such a way that it is impossible to tell the difference than is that still agency?
Technically no, but we wouldn't be able to know otherwise. That gap is closing.
No there is a logical errror in there. You are implicitly asserting that the trained thing is an imitation, whereas it is only the output that is being imitated.
A flip way of saying it is that we are evolving a process that exhibits the signs of what we call thinking. Why should we not say it is actually thinking?
How certain are you that in your brain there isn’t a process very similar?
I am simply asking a question. If anything I am only asserting the possibility that it is an imitation. I am more saying that there is no method to tell the difference on which possibility is true. Is it an imitation or is it not? The argument is ultimately pointless because you cannot prove it either way.
The only logical error is your assumptions and misinterpretation of what I said and meant.
But to carry your argument one step further, if there is no difference between imitation and the real thing, is there anything meaningful to be debated here? "Is it an imitation or is it not?" isn't even a valid question in that context. Imitation === The Real Thing.
I literally told you what I was asserting and made it completely explicit. So what you assumed I was implying was wrong.
I never said there is no difference. There is a difference, the difference is just not discernible or observable.
Let me give you an example. It’s like an unsolved murder. You find a victim who is stabbed, you know he was killed, we know someone killed him, but we don’t know who.
In the case of AI is the same. We know certain things about it, but if it produces output indistinguishable from AGI then we cannot discern whether it is an imitation or the actual thing. There does exist a difference but we cannot meaningfully determine it either way in the same way we can’t solve an unsolvable murder. But just because we can’t solve a murder does not mean there was no perpetrator.
reply