It's also a context-specific scale. I work in computer vision. Building the surrounding app, UI, checkout flow, etcetera is easily Level 6/7(sorry...) on this scale.

Building the rendering pipeline, algorithms, maths, I've turned off even level 2. It is just more of a distraction than it's worth for that deep state of focus.

So I imagine at least some of the disconnect comes from the area people work in and its novelty or complexity.

This is exactly true in my experience! The usefulness of AI varies wildly depending on the complexity, correctness-requirements, & especially novelty of the domain.

This attribute plus a bit of human tribalism, social echo-chambering, & some motivated reasoning by people with a horse in the race, easily explains the discord I see in rhetoric around AI.

am layman. is CV "solved" at this point, or is there more work to be done?

Far from solved! Though, like seemingly everything, it has benefited from the transformer architecture. And computer vision is kind of the "input", it usually sits intersecting with some other field i.e. cv for medical analysis is different to self driving is different to reconstruction for games/movies.

But why are you making projects in so many languages? The language is very rarely the barrier to performance, especially if you don't even understand the language.

I try to pick the language best to the situation rather than giving into my own biases. I need to broaden my horizon to be able to cover the full stack of stuff that I need, not just the things I've been doing myself a lot for years. There's a lot of stuff that used to be out of my comfort zone that I can now tackle easily. Stepping over my own biases is part of that.

I know not everybody is quite ready for this yet. But I'm working from the point of view that I won't be manually programming much professionally anymore.

So, I now pick stuff I know AIs supposedly do well (like Go) with good solid tool and library ecosystems. I can read it well enough; it's not a hard language and I've seen plenty other languages. But I'm clearly not going to micro manage a Go code base any time soon. The first time I did this, it was an experiment. I wanted to see how far I could push the notion. I actually gave it some thought and then I realized that if I was going to do this manually I would pick what I always pick. But I just wasn't planning to do this manually and it wasn't optimal for the situation. It just wasn't a valid choice anymore.

Then I repeated the experiment again on a bigger thing and I found that I could have a high level discussion about architectural choices well enough that it did not really slow me down much. The opposite actually. I just ask critical questions. I try to make sure to stick with mainstream stuff and not get boxed into unnecessary complexity. A few decades in this industry has given me a nose for that.

My lack of familiarity with the code base is so far not proving to be any issue. Early days, I know. But I'm generating an order of magnitude more code than I'll ever be able to review already and this is only going to escalate from here on. I don't see a reason for me to slow down. To be effective, I need to engineer at a macro level. I simply can't afford to micro manage code bases anymore. That means orchestrating good guard rails, tests, specifications, etc. and making sure those cover everything I care about. Precisely because I don't want to have to open an editor and start fixing things manually.

As for Rust, that was me not thinking about my prompt too hard and it had implemented something half decent by the time I realized so I just went with it. To be clear, this one is just a side project. So, I let it go (out of curiosity) and it seems to be fine as well. Apparently, I can do Rust now too. It's actually not a bad choice objectively and so far so good. The thing is, I can change my mind and redo the whole thing from scratch and it would not be that expensive if I had to.

In my experience Rust and Go are both opinionated languages with strong types which makes them work well with agentic coding.

Yes because in most contexts it has seen "caveman" talk the conversations haven't been about rigorously explained maths/science/computing/etc... so it is less likely to predict that output.

Do you have any evidence at all of this? I know how LLMs are trained and this makes no sense to me. Otherwise you'd just put filler words in every input

e.g. instead of: "The square root of 256 is" you'd enter "errr The er square um root errr of 256 errr is" and it would miraculously get better? The model can't differentiate between words you entered and words it generated its self...

It's why it starts with "You're absolutely right!" It's not to flatter the user. It's a cheap way to guide the response in a space where it's utilizing the correction.

What do you think chain of thought reasoning is doing exactly?

People have researched pause tokens for this exact reason.

You’re conflating training and inference

If only there were some system where the incentives could freely flow through and permeate every level of the sector. Where those organisations that provide sub-standard care die and those that excel receive outsized funding...

Unfortunately, a system with these qualities doesn't exist in practice. You just end up with the same too-big-to-fail macro organization minimizing their point-of-care labor spend and maximizing their management spend either way.

I first learned when reading about Steve Jobs, how the Japanese never use "quality" in their advertising. Yet people still view(ed, at least) Japanese manufactured goods as superior quality. It turns out people don't judge quality based on what you tell them but based on their experience.

All this to say: I wouldn't stress about it too much. In the consumer space the best usually does win, and people will simply vote with their feet.

Yeah, well, that's how capitalism was meant to work, but mostly seems to have been implemented with various thumbs on scales (which is also true for most of the -isms).

Feels kinda like the thumbs on the scales resulted in the (d)evolution towards techno-feudalism.

Gotta give it to the Cursor team, they must have REALLY good numbers. They raised at a 9.9b valuation less than a year ago and now apparently targeting 50b.

Makes no sense to me, the main driver of codex, Claude code, etc.. seems to be fixed cost plans that offer reduced token cost. Cursor doesn’t have a good model so they can’t offer that (at least not to the same extent).

Composer 2 is great tbh. It makes my over runs of the ultra plan much less painful.

But why? We’ve not colonised either of the poles of our own planet in any real way out of a sense of prestige. Heck there’s huge areas in Canada and Russia uninhabited and these are all a dream to live in compared to Mars.

Turns out the real overpopulation is in places people want to live.

Because I don't believe our species should be trapped this planet forever. If we don't become multiplanetary now, then when? And there is an incredibly short window for us to become multiplanetary. We currently live in a golden era of abundance that will not last, and we must make the most of this time period.

I think most people don't realize how inherently unstable our society is, and how quickly civilization can devolve.

Nuclear war is a huge issue. We've had three conflicts this decade that could have led to a nuclear war. All of which are still unresolved.

> Nuclear war is a huge issue. We've had three conflicts this decade that could have led to a nuclear war. All of which are still unresolved.

I'm kicking myself for engaging with this at all, but that's poor reasoning if you're worried about nuclear war. The risk of MAD forces a detente, if there were a (perceived) hedge against it, that increases the likelihood of MAD happening.

If nuclear war happens, Mars colonies depend on expensive, technical supply chains on Earth that will be destroyed.

We take for granted a whole damn planet where water falls from the sky, food and fuel come out of the ground and there's abundant amount of replenishing atmospheric O2 available for ubiquitous reaction and combustion.

Without resupply from a nuked Earth, you're left with the fact that food, manufacturing, construction, etc all depend on, at the very least, atmospheric oxygen, and Mars will never hold a meaningful atmosphere. Without atmospheric oxygen, and thus combustion, when it comes to supply chains required for existing anywhere, you aren't building infrastructure, you aren't growing food without nutrient supplies, and you aren't manufacturing sustainably, efficiently, or at all.

And that ignores that Martian dust and soil is toxic[1] to life, which requires even more resources to mitigate, remove, keep out/off of people and living things, and even more resources to treat and maintain the soil if you ever want to use it to grow food.

Earth is the one shot people have, and a nuked Earth is infinitely more habitable than Mars. Even the bottom of the ocean is more habitable than Mars. It just does not make sense as a backup option to Earth. And if Mars is a pipe dream, life isn't leaving this solar system and surviving independently as anything resembling humans.

[1] https://en.wikipedia.org/wiki/Martian_regolith#Toxicity

I think the prestige, overpopulation, and pollution arguments all suck. The important differences are that the poles are not political free-for-alls that people can just colonize, and everything is still 1g vs. Mars' 0.38g.

Redundancy is the answer.

Amazon Leo just signed delta as a customer so competition is indeed close behind.

I think SpaceX is an incredible company but at this valuation I’d expect it to have something as pervasive as the iPhone or Nvidia chips. It seems to have only small niches.

"Hire" an unfortunately ironic typo!