Over-reliance on AI tools is probably going to be seen as one of the great mistakes of this generation of engineers.
I use AI to work through certain problems or help with semantics. 99% of the actual work still comes from my experience and skill.
My workflow is the same workflow I've always used. Work until I get stuck, have a five minute conversation with a colleage who may be slightly drunk, extract any useful information, then attempt to integrate that while I get back to work.
Yes, but I don't see it as problematic. When I'm in flow, I don't need the AI. I'm working on something I grok and I can see through from start to finish. Achieving flow on something you don't grok is a nonsesnse question. When I need the AI for something, it's either something I'm too tired for, something that doesn't require thought, or it's something that I need to stop and think through. All of which explicitly require exiting flow to wrangle with tooling and research.
If AI breaking flow is a problem for you, I'd suggest you're doing AI or flow incorrectly. Don't ever expect that AI will do anything other than disrupt you. Just ask yourself: would you be comfortable snapping off this task to a green intern while you stay in flow? Or would they need some level of your attention and supervision to not set the shop on fire? AI is not a tool to keep you in flow, it's a tool to help shorten the downtime between flow states.
I only ever use AI tools when flow state is light, or flow is hard blocked by something I can't immediately resolve on my own. Using AI tools in the middle of my flow sounds stupendously disruptive and counter-productive.
My understanding of the delayed choice experiment is that causality is preserved only because you can't actually determine the result of a single event without comparing the two results. It's spooky action at a distance, except it's temporal instead of spatial distance.
Which honestly I find to be pretty flimsy reasoning. It's almost ontological: causality is preserved because we can't prove that it isn't.
I know it's more complex than that, but it still feels like papering over a hole in our theories.
Did we collectively forget that most written languages directly encode the sounds of the spoken language?
Your brain tokenizes sounds into words. A beginner reader has to parse text into sounds and then into the token. An advanced reader can skip the middle step and parse text into tokens. But you still have to know how to parse text into sounds, there's no way around it.
It'd be like giving someone a French texbook, only instruct them in English, don't even mention the different sounds, and somehow expect them to learn conversational spoken French. It's nonsense.
They are pretty common in Europe, because older houses here all had gas-powered furnaces, and with an air-to-water heat pump you can continue to use the existing tubing and radiators.
New house builds often include one to run domestic hot water and under-floor heating for bedrooms and such. The downside here is that you probably also need an air-to-air heat pump for cooling in summer, and now you have two expensive heat pumps, and a whole extra set of pipes running around...
Honestly at this point who would seriously use any Microsoft UI framework? They've abandoned 100% of their previous UI frameworks unfinished when they get distracted by a new, shinier framework.
Why use a busted incomplete framework missing basic features when there's entire ecosystems of open source cross-platform frameworks being actively maintained and which actually have all the features you need?
Really this is just another UWP destined to be forgotten and scorned.
WPF uses DirectX 9, in any case it is good enough for many businesses use cases.
UWP never had feature parity with neither Forms nor WPF, so already it has a hard sell to businesses.
Microsoft marketing always cool about cool experiences and design, without the actual meat in features.
That is why there were API reboots between Windows 8, Windows 8.1 and Windows 10.
Then comes along Project Reunion, which is supposed to properly unify Win32 and UWP execution models, instead a few months in, they do yet another reboot, that to this day is a shadow of UWP features, endless bugs, and a team that everyone thought was also part of the layoffs, has they were radio silence for months.
Forms is still my go-to for quick one-off and disposable programs. But maily because it's simple and I have a big library of controls and extensions I've built over years to make Forms more complete.
I've also tried WPF, but it's missing very basic controls that I just don't feel like recreating.
I give coding tests on real(ish) problems adjacent to the job. It's one half showing me what you can do, and one half seeing how you deal with a real work environment, collaborating with others, unknown requirements.
It's not a puzzle, it's not a trick question. It's simply a matter of 1) can you do the job and 2) do you get along with the team.
The last interview I did, we had the candidate build a game of marbles in Unity. None of us knew the rules, so it turned into a collaborative process between the candidate and our engineers. We learned that the candidate can think on his feet and fluently adapt his program to requirements as they manifest. Plus it was just a lot of fun. He was one of our best hires, too.
That's the real problem with the current process. It assumes all the candidates are just random people without any credentials. So candidates need to prove that they can code each and every time. Even a Phd degree doesn't count. Years of work experience doesn't count. And they test you with challenging competitive programming questions. Not even just normal programming. This process used to be only for Google etc, where demand is too high. It's just an elimination process. But today every company in every country blindly copies this insanity. Solving a hard dynamic programming question in 20 minutes after 2 months of studying means nothing from a real engineering perspective.
Exactly. There's no FizzBuzz for a doctor, because the medical degree, residency, and passing the medical board exam provide enough of a signal of basic competence.
In Germany that would be a recomendation letter, pretty much valuable by most HR departments, and not showing up with one isn't really recommended, a shitty one is better than none at all.
Yea apprenticeship certification is what I have (Germany). It's a lot different from other countries apprenticeships and shows ppl "Hey that dude worked this job for 3 years".
College degree should also work as you usually gather some job experience during university.
What i not meant were specific certificates like the Cisco ones.
Even a letter of recommendation from a previous employer would count for me.
I use AI to work through certain problems or help with semantics. 99% of the actual work still comes from my experience and skill.
My workflow is the same workflow I've always used. Work until I get stuck, have a five minute conversation with a colleage who may be slightly drunk, extract any useful information, then attempt to integrate that while I get back to work.
It's a tool to solve problems, not do work.