When we have AGI, we'll have self-driving cars. We aren't getting either in a year's time. The need for white-collar jobs in areas will shrink (not disappear), possibly to expand elsewhere.
Yep, really advanced Google searches were never that good. LLM, yeah, it halucinates, it's never spot on but as sure as heck it knows what I'm trying to ask. It doesn't give me arborists if I say something like "list tree searches".
I always assumed what I felt was obvious: Numbers 21:4-9, where God instructs Moses to make a bronze serpent and place it on a pole to heal Israelites dying from poisonous snake bites.
The Bible contains many verifiable historical references supported by archaeology and ancient records, even as a primarily theological text. Here are three examples:
1. The Tel Dan Stele and the House of David
This 9th-century BCE inscription records an Aramean king’s victories over the “king of Israel” and the “king of the House of David.” It gives the earliest extra-biblical proof of the Davidic dynasty at the heart of the Hebrew Bible.
2. Sennacherib’s Prism and Hezekiah
The prism names “Hezekiah the Judahite,” details the capture of 46 cities, and describes besieging Jerusalem in 701 BCE, closely matching 2 Kings 18–19, including the tribute paid.
3. The Pilate Stone and Pontius Pilate
Found at Caesarea Maritima, this inscription names “Pontius Pilate, Prefect of Judea” under Tiberius, directly confirming the Roman official who tried Jesus in the Gospels.
These independent sources from rival powers strongly support the Bible’s historical value where evidence exists.
The logical conclusion is to use "a Haskell" typed language that will ensure every path is considered to guard against AI mistakes. OTOH, clojure repl, expressibility, immutability, and data-driven nature has its own advantages. Tacking on malli (runtime type checking) or spec (types/contracts) helps LLMs avoid type problems altogether or at least confront problems during testing.
For core system functionality maybe. But for most applications Rust slow compiler iteration speed becomes a bottleneck when the likes of TypeScript (with Bun) and Go have sub second iteration times.
Plus AI is also good at catching, in other languages, errors that Rust tooling enforces. Like race conditions, use after free, buffer overflows, lifetimes, etc.
So maybe AI will become to ultimate "rust checker" for any language.
In my experience developing different types of applications in Rust, the claims of a "slow compiler" are overstated. Sub second iteration times are definitely a thing in Rust as well, unless you're adding a new dependency for the first time or building fresh.
Our experiences clearly differ then. And for others as well since it's a common complain.
Countless time I have seen other people complain as well. There are articles about it even. Can't find the YouTube link now but recently a gamedev abandoned Rust due to compilation speed alone because iteration speed was paramount to their creative process.
Handwaving isn't going to make it any better. And thinking Go/TS compilation speed are comparable to Rust is, a handwave and a half to say the least.
Cargo check and friends are subpar for AI because they actually need to run the thing and unit tests for efficient agentic loops.
A single loop might recompile and rerun the application/unit tests enough times that slow compilers like Rust and Scala become detrimental.
I think you could have left it at differing experiences and not gone further saying I'm handwaving anything. That doesn't seem productive.
I'm not saying that Rust compilation time is comparable to Go/TS, I'm saying the blanket claim that Rust iteration speed will be a bottleneck requires context.
I definitely agree with you that it is a complaint that is often repeated online, but that doesn't make it universally true. In my experience it's a claim that is often echoed without proper context.
Particularly in the case of AI Rust recompliation times in my experience have not been the dominant cost, but are instead overshadowed by inference time, the agent working through different approaches, etc.
The productivity increase I get overall by not having to worry so much about if my rust code will work if it compiles tends to net faster iteration speeds for me. Compile times have never bothered me.
I'll propose this as the only unbreakable law: "everything in moderation", which I feel implies any law is breakable, which now this is sounding like the barber's paradox. What else does anyone propose as unbreakable?
Saying this is like saying 'pick the optimum point' without saying anything about how to find the optimum point. This cannot be a law, it is the definition of optimum.
Note that optimum point need not be somewhere in the middle or 'inside', like a maxima. The optimum point could very well be on an extreme of the domain (input variables space).
It's easy to chalk it up to "fear of the unknown", when in reality it's both good and bad depending on who's wielding it. It can be used to tear down or build up, solve problem or create problems just like every advance before it. So while I'm generally excited with where it can go, I guess I don't mind being reminded there can be downsides.
Anecdotally, I was just there and ran into a couple of anti-AI people and it's not like I was bringing it up. All I got was they were worried about the water and the heat produced. I wonder if someone has done an analysis of old-search vs AI-search, I most definitely get the info quicker that I want with AI, does that make up for the LLM cost, I have no idea.
reply