That sounds like very much like the reviews I used to get on papers at top tier conferences before the advent of LLMs.

It's rare I have to do anything client side, but I'm so grateful for Clojurescript when I do. I used to dread native Javascript and the crazy tooling, but Clojurescript makes building client side functionality an absolute joy. It's just a shame it's not all that popular

Unfortunately, there isn't demand and there's no such thing as a high demand tech professional anymore. The EU is way over saturated with tech talent and that's unlikely to change.

This have implications for public key cryptography?

My naive notion on this is yes, iff the new method is computationally or memory-wise of lower complexity

Computing M_a(n) appears to be at least as hard as factoring n for a=1, so I think you're safe here.

My frustratingly rubbish M1 Air does this, amongst many other odd and annoying things. But hey... what are the alternatives?

Although most terminals (stty)/UI wrappers/multiplexers etc. would allow this as a configuration, I believe ElementaryOS and others are deliberately setup out the box to offer a Windows-like experience, including preserving keyboard shorts, without further configuration if this is what you seek.

There's not really an issue with planning. The actual problem is that developers will then sit on that land rather than build on it, as this is more valuable long term. See "land banks"

I'm baffled how we've come to accept such poor quality software in our products and key services. We seem to shrug off software issues that would deter purchase if they were material or mechanical.

How is this a software quality issue? This is a process issue with the rental company. It doesn't sound like the car itself is doing anything wrong.

I don't have a BMW, so I may be wrong, but from the comments it sounds as though:

1) physical access gives remote privileged access - this is a car, not a server! We don't lock cars in data centers, friends, family, mechanics, the car wash, valets or an entire custom base in this case may have access.

2) there is an idiosyncratic process that must be followed to ensure the vehicle is suitable for this use which is sufficiently obscure that it's often overlooked. The risks of which, are quite grave (remote tracking).

3) even if everything is done "correctly", it will limit the functionality to the product you've paid to rent.

If you were the owner of such a car, how exactly would you expect to link your phone to the car using the first party app if not for physical access?

There's a standard reset process that the rental company is simply not following here

I built an exploding pie, put it in the fridge with a "do not eat me" sign, not my problem somebody blew off their arm.

Processes should/must be built with the users in mind. Users do unexpected, stupid things.

So, yeah, the car itself isn't doing anything wrong, but the whole "fleet use" system designed by BMW+fleet owners appears massively flawed. And that system includes all the software.

> the whole "fleet use" system designed by BMW+fleet owners appears massively flawed.

It would seem that the system works just fine, it's just completely being ignored by the rental company.

I could just as easily tape a cheap cell phone inside a hotel room to the wall and record the other guests with it. If the hotel cleaning staff doesn't clean the room and remove the device, how is that any different?

Really?

I just wanted to second the previous comment, and this is even for adjacent fields. Also a PhD AI/ML grad, and so many of us are out of work at the moment that we'll happily settle for prompt engineering roles, let alone RAG etc., just to maintain appearances on CVs/eligibilty for possible future roles.

Kinda surprised of that, actually. Sure, I get that research interest in any if the "traditional" ML methods (SVMs, markov models, decision trees, that kind of stuff) is probably essentially dead right now, but I had thought interest in neural networks and "understanding" what LLMs do internally to be ballooning.

I could imagine that even those "ancient" techniques might some day make a comeback. They're far inferior to LLMs in terms of expressive power, but they also require literally orders of magnitude less memory and computation power. So when the hype dies down, interest in solutions that don't require millions in hardware cost or making your entire business dependent on what Sam Altman and Donald Trump had for breakfast might have a resurgence. Also, interestingly enough, LLMs could even help in this area: Most of those old techniques require an abundance of labeled training data, which was always hard to achieve in practice. However, LLMs are great at either labeling existing data or generating new synthetic data that those systems could train on.