When it comes to full on vibe coding (Claude Code with accept all edits), my criteria is whether I will be held responsible for the complexity introduced by the code. When I've been commissioned to write backend APIs, "the buck stops with me" and I will have to be able to personally explain potentially any architectural decision to technical people. On the other hand, for a "demo-only" NextJS web app that I was hired to do for non-technical people (meaning they won't ever look at the code), I can fully vibe code it. I don't even want to know what the complexity and decisions AI has made for me but as far as I am concerned this will be a secret forever.
The ultralytics/ultralytics repo is pretty beginner friendly (kudos to them for that) but I surmise that it therefore draws a lot of beginner level coders who can't immediately tell that the AI generated "solutions" are bs.
I've made several contributions to their main repo and the LLM generated mush replies from various core team accounts have been a horror, derailing Issues threads and such. An excellent case study in how not to use LLMs.
On the whole I agree, but having read "Reinforcement Learning: An Introduction" by him and Barto, this article comes across as a hardly nuanced self-endorsement of RL as "the inevitable future of AI". Without mentioning RL by name (but hinting at it with HMM, search&learning- like exploration&exploitation), I think he might be suggesting that any supervised learning is still too specific. Not to mention that he works for DeepMind, which has found fruitful applications of RL.
Proponents of unlimited speed Autobahn could argue: 1. The tourism value: I anecdotally know people visit Germany just for the Autobahn. (See Nürburgring, which is technically an Autobahn) 2. Sale of performance cars: The high actual minimum speed (from my experience you need to drive at around 150km/h in the middle lane, and 180-200km/h to pass on the left lane). Almost anywhere else in the world, the top speed of cars is irrelevant. In Germany, however, high performance cars (inevitably, German brand ones) can be desired for their better handling at speeds at or above 200km/h. 3. Germans collectively have excellent driving mannerisms and skills which were a result of unlimited speed. It is probably true that enforcing a speed limit today would decrease deaths, but the next generation would then regress to the mean.
I would be interested to see someone quantify the above points and compare them with the economic costs discussed in the OP article.
Another significant but unnoted design feature is the ability to pass a strap through multiple jerrycans in a row. Essential for mounting them externally to Humvees and MATVs.