I will be the first to acknowledge that humans are a bad judge of performance and that some of the allegations are likely just hallucinations...
But... Are you really going to completely rely on benchmarks that have time and time again be shown to be gamed as the complete story?
My take: It is pretty clear that the capacity crunch is real and the changes they made to effort are in part to reduce that. It likely changed the experience for users.
IMO, this isn't entirely a "new world" either, it is just a new domain where the conversation amplifies the opinions even more (weird how that is happening in a lot of places)
What I mean by that: you had compiled vs interpreted languages, you had types vs untyped, testing strategies, all that, at least in some part, was a conversation about the tradeoffs between moving fast/shipping and maintainability.
But it isn't just tech, it is also in methodologies and the words use, from "build fast and break things" and "yagni" to "design patterns" and "abstractions"
As you say, it is a different viewpoint... but my biggest concern with where are as industry is that these are not just "equally valid" viewpoints of how to build software... it is quite literally different stages of software, that, AFAICT, pretty much all successful software has to go through.
Much of my career has been spent in teams at companies with products that are undergoing the transition from "hip app built by scrappy team" to "profitable, reliable software" and it is painful. Going from something where you have 5 people who know all the ins and outs and can fix serious bugs or ship features in a few days to something that has easy clean boundaries to scale to 100 engineers of a wide range of familiarities with the tech, the problem domain, skill levels, and opinions is just really hard. I am not convinced yet that AI will solve the problem, and I am also unsure it doesn't risk making it worse (at least in the short term)
I genuinely think systems like Unison are "the future of computing"...
But the question is when that future will be.
Part of the beauty of these sorts of systems is just that the context of what your system actually does is in one system, you aren't dealing with infra, data, and multi-service layers
Maybe that means it is a much better foundation for AI coding agents to work in? Or maybe AI slows it down? we continue to try and throw more code at the problem instead of re-examining the intermediate layers of abstraction?
I really don't know, but I do really want to learn more about is how the unison team is getting this out in the market. I do think that projects like this are best done outside of a VC backed model... but you do eventually need something sustainable, so curious how the team things about it. Transparently, I would love to work on a big bet like this... but it is hard to know if I could have it make financial sense.
With all that, a huge congrats to the team. This is a truly long-term effort and I love that.
not disagreeing with your point here, or in the follow-ups of the pain of https for "local network" apps... but I really wish that we could get to a place where we could get away from this distinction. Obviously, ipv6 is not that easy or realistic, but that really is, imho, the "right" long term answer.
Having gone down the path of being able to just spin up "local" services that get a publicly routable (but most often firewalled off) ipv6 IPs and then good DNS integration is really neat... but still requires lots of technical chops. I wish that weren't the case
My point is not that value extraction wouldn't happen, my point is simply that in addition to the value extraction we also made other huge shifts in economic policy that taken together really seem to put us on a path towards an "AGI or bust" situation in the future.
Is that a bit hyperbolic? isn't this just the same as dotcom and housing bubbles before where we pivoted a bit too hard into a specific industry? maybe... but I also am not sure it would be wise to assume past results will indicate future returns with this one.
> I was discussing with a friend that my biggest concern with AI right now is not that it isn't capable of doing things... but that we switched from research/academic mode to full value extraction so fast that we are way out over our skis in terms of what is being promised, which, in the realm of exciting new field of academic research is pretty low-stakes all things considered... to being terrifying when we bet policy and economics on it.
That isn't overly prescient or anything... it feels like the alarm bells started a while ago... but wow the absolute "all in" of the bet is really starting to feel like there is no backup. With the cessation of EVs tax credits, the slowdown in infra spending, healthcare subsidies, etc, the portfolio of investment feels much less diverse...
Especially compared to China, which has bets in so many verticals, battery tech, EVs, solar, then of course all the AI/chips/fabs. That isn't to say I don't think there are huge risks for China... but geez does it feel like the setup for a big shift in economic power especially with change in US foreign policy.
I think the answer to that right now is highly workload dependent. From what I have seen, it is improving rapidly, but still very early days for the software stack compared to Nvidia
But... Are you really going to completely rely on benchmarks that have time and time again be shown to be gamed as the complete story?
My take: It is pretty clear that the capacity crunch is real and the changes they made to effort are in part to reduce that. It likely changed the experience for users.