Scea91

Why a multiplayer game when you can do the same test in a nicer environment at a local golf course?

I used Claude Code in tmux since January until August.

Now I am 100% in the ChatGPT desktop app, because it felt I am doing way too much tinkering amd maintenance of my setup instead of doing real work.

Retrospectively, it feels tmux was right choice early this year as the alternatives just weren’t there. It was also fascinating and enjoyable building own workflow from the primitives and I did learn a lot, but the learning has plateued and what was left was mostly chore.

It feels there is a bit of contradiction in your post. You posit love, struggle and basically the current human experience as meaningless and yet you pitch biological immortality.

How would we spend this immortality? Seems good to have this sorted out before gambling everything on it.

The civilian cost was intentionally not part of the equation. The toxic effects were known and the numbers show that it was not a few isolated accidents but a intentional disregard for the human cost.

2) Unintentional side-effect affecting millions of Vietnamese people? Hundreds od thousands of children born with birth defects.

The wording seems to confirm it was part of training data at some stage. If not, they would be able to prove it quite easily I assume.

Same can be said about human review if the argument is non-determinism.

Its a problem of the individual institutions and countries. The budget required for AI tools currently is negligible compared to other university expenses. We don't need to call everything a systemic disadvantage when the disadvantaged (at the institution level) have agency here.

I don't think number of parallel agents is the right productivity metric, or at least you need to account for agent efficiency.

Imagine a superhuman agent who does not need to run in endless loops. It could generate 100k line code-base in a few minutes or solve smaller features in seconds.

In a way, the inefficiency is what leads people to parallelism. There is only room for it because the agents are slow, perhaps the more inefficient and slower the individual agents are, the more parallel we can be.

> Parents can’t help but worry about how a lack of AI preparedness will affect their kids’ future career prospects.

The parents I've interacted with who have this worry tend to be the least tech/AI literate. It reminds me of the pressure some parents in my non-English speaking country place on kids and English teachers. Usually the parents who don't understand English well themselves and are full of anxiety about it.

Realistically, it's been just over 3 years since ChatGPT arrived and the paradigms shift every year. All of us had to learn on the go as adults, in a short time, and we did. So why worry so much about kids missing out?

I am afraid that in a total war, all available resources will be used.

I am pretty low confidence here but I think in theory it should but in practice there is no mechanism enforcing that?

UMA's current market cap is $68M. There are some Polymarket markets far exceeding that.

UMA's security model assumes the cost to corrupt the oracle exceeds the profit from corruption. It is quite interesting because it doesn't consider the Polymarket side at all in the calculation.

Doesn't this whole model break down when the Polymarket market far exceeds UMA's market cap?

> - (~2023) Ok, it can write a full function, but it can't write a full feature.

The trend is definitely here, but even today, heavily depends on the feature.

While extra useful, it requires intense iteration and human insight for > 90% of our backlog. We develop a cybersecurity product.