> You don't direct the action of an LLM - you make requests and it acts on its own, without your energy or continuous direction. This request/automation cycle is not wielding, it's not control, it's not continuous, it's without feedback.
There are certainly many ways of using coding agents, but the way he describes it is not my approach. There's a huge amount of direction and feedback.
I don't know if I'd call it a tool so much as a modality of building software. Within the modality, you have everything from truly unsupervised YOLO coding to something far more structured. To an extent, it's similar to how you go about setting up standard operating procedures to get teams of humans to do work at scale.
I'm really sorry you have gone through that. I have been pretty deeply depressed a couple times in past, to the point of considering suicide. What I share with people is that in those periods, it truly felt like things weren't going to get better and I was just going to be in emotional pain indefinitely.
But gradually, things got better in my life. I have had so much joy in the 20 years since. I still have bad days sometimes, but on the whole, I feel such a sense of gratitude.
I can't promise anyone that their own situation will get better. I can only say that my experience is that it is possible.
I used to use a TI-89 in high school, which had a calculus-enabled CAS built-in. It was really handy for checking my work and for working on components of problems. And if you're in a class where the work is more applied math, like physics, it's really helpful so that you can think of the overall problem without getting bogged down in the computation.
> Decision models have the potential to have an even larger impact on the Real World than LLMs have to this point (which is obviously quite large). But the model itself matters less than the product experiences you build around the model, and its very likely that the incumbent labs are treating the area as something more like "oh yeah I guess we can ship that and then forget about it" rather than investing in what building business processes on decision models looks like.
I agree that the potential is enormous, but I'm not sure why it would be Typesafe who figures out the best way to develop and package this capability. They have a bit of an advantage in the amount of time they have been thinking about this problem specifically, but I think it's a big leap to say that they will be able to dominate the category.
> Which is why these countries are so much more prosperous when outsiders rule them, supressing the clan systems for anything more meritocratic.
That's generally not how colonization worked. Quite often, colonization intentionally created the large scale tribal identities. Allegiance to clans grows because of deep suspicion of extractive governments put in place by outsiders.
The colonial system was meritocratic. It has a lot of bad things, but an absolutely staggering amount of Englishmen and Scotsmen and Welshman got their start and career through their hard work in India. A lot of whom, especially those minorities, would never have had a chance in G.B.
- Heart of Darkness on British imperialism in the Congo
The cost of the socio-economic mobility provided to lower classes in colonizing societies was absolutely staggering. Unimaginable cruelty, suffering and destruction was visited upon the native populations.
Europe didn't exist as a cultural concept at that time. There was Rome, but that was a Mediterranean empire with hugely significant southern European, east Asian, and north African components. Citizens and elites came from all parts.
Eventually, we get a Europe united under Christianity, but I wouldn't call that fully developed until the end of the Viking age.
Okay, so if they don't police themselves, then they're full of crap. But then what?
David Sacks holds an advisory role in the government. If all of this stuff goes belly up and all he was doing was tweeting, then he is not only full of crap, but failing the citizens he's supposed to be serving.
The book Seveneves is about civilization confronting impending doom from the moon exploding. I never finished the book because I felt it was depressingly realistic.
There are certainly many ways of using coding agents, but the way he describes it is not my approach. There's a huge amount of direction and feedback.
I don't know if I'd call it a tool so much as a modality of building software. Within the modality, you have everything from truly unsupervised YOLO coding to something far more structured. To an extent, it's similar to how you go about setting up standard operating procedures to get teams of humans to do work at scale.
reply