Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
ResearchCode
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
31.
▲
by
ResearchCode
3y ago
Learn the concept better, not an exception to an incorrect rule. It seems like the model architecture just isn't good enough.
32.
▲
by
ResearchCode
3y ago
LLM engineers rushing to hard code B is A in their "AI". They call it reinforcement learning with human feedback.
33.
▲
by
ResearchCode
3y ago
You need business with them to browse the internet and now to even buy a car. There is a word for this.
34.
▲
by
ResearchCode
3y ago
You don't seem to understand them. German law does not force you to have a three month notice period at the start of your employment, you can negotiate for a shorter period without the at-will part.
35.
▲
by
ResearchCode
3y ago
That's very courteous, 28 days is plenty of time to find a new tech job. A real courtesy would be a golden parachute worth at least 6 months wages if you get laid off without cause. That's probably not more than a US sign-on bonus
36.
▲
by
ResearchCode
3y ago
Or you don't and end up making a non-functional website for the public sector for $100M. Software projects fail often, and not for the lack of trying, there is not yet any evidence that you can micromanage them into succeeding.
37.
▲
by
ResearchCode
3y ago
It's pretty bold to give a trial period to an international hire. Employment security is a good thing, and since wages are much lower than in the US, there is not much risk.
38.
▲
by
ResearchCode
3y ago
I don't see the risk. It seems German companies are making weak offers with "probation periods" even for international hires. One problem might be that such offers are not very attractive.
39.
▲
by
ResearchCode
3y ago
Get rid of the scheduling too if you're doing it too often. Two times a year is good enough.
40.
▲
by
ResearchCode
3y ago
You're not wrong. Applied ML articles are not worth reading.
41.
▲
by
ResearchCode
3y ago
They don't help you do that, even though poor non-technical managers might think they do. That's why good software projects don't use them. I don't see any story ticket velocity points used in Linux kernel development. Y
42.
▲
by
ResearchCode
3y ago
Then get rid of the "points" and maybe the "tickets" too.
43.
▲
by
ResearchCode
3y ago
I don't think they did. If churning out low-quality code for 6 months was good project management, the Linux Foundation would be doing it.
44.
▲
by
ResearchCode
3y ago
Those are not important at all, or the top software projects would be using them. Linux kernel is doing fine without the velocity point story tickets. Agile "planning" which is used in worse software projects is snake oil.
45.
▲
by
ResearchCode
3y ago
It is, unless you spend too much time in a lecture format with too little deep practice. Taking work home is not a good solution.
46.
▲
by
ResearchCode
3y ago
Practice during school hours is more than enough. Primary education is not so demanding that you need more to succeed.
47.
▲
by
ResearchCode
3y ago
The measure is called an accepted pull request. You don't need a ticket to submit a patch to the Linux kernel. If you're in a dysfunctional agile micro-management environment with "stories" and "backlogs" then
48.
▲
by
ResearchCode
3y ago
It's not a problem, that's how you create software. You can put some of those initiatives on a "road map" if you want, but there must be space for them. 50% "slack time" is a good standard for software engineer
49.
▲
by
ResearchCode
3y ago
The "intelligence" of large language models needs to be evaluated like the abilities of self-proclaimed psychics. You send your binary to an independent third party and who evaluates it on new problems. It's only a "Huma
50.
▲
by
ResearchCode
3y ago
We are already five years late in autonomous vehicles replacing all truck drivers. We will see if we even have autonomous driving "long before" 33 years have passed. "AGI" (rebranded AI after "AI" failed to d
51.
▲
by
ResearchCode
3y ago
Come close to solving the toy problem of autonomous driving first, we're still waiting.
52.
▲
by
ResearchCode
3y ago
Join a team that doesn't do "sprints" if you want to do meaningful work.
53.
▲
by
ResearchCode
3y ago
It can be reduced, like by cancelling most meetings that could have been an e-mail.
54.
▲
by
ResearchCode
3y ago
It's better to go somewhere where your colleagues and hopefully management are intelligent enough to recognize loud laboring as vulgar and an anti-signal.
55.
▲
by
ResearchCode
3y ago
If you look at the valuation of companies that hire or were started by top talent, all companies should want them. Crab mentality can make you rationalize scaring them away.
56.
▲
by
ResearchCode
3y ago
There's an abundance of people that can talk but do less. Real tech talent are rare, not that you find them by algorithms quizzes.
57.
▲
by
ResearchCode
3y ago
If you're doing a PhD, that point is a weekly chat with your supervisor. I don't see why a few hour "rabbit hole" would be a problem in software engineering. If we're even discussing hours unless billing hourly, tha
58.
▲
by
ResearchCode
3y ago
Those "rabbit holes" are where you learn. Saving hours is more important if you bill hourly, less if you are salaried.
59.
▲
by
ResearchCode
3y ago
That sounds like micromanagement. Are you a contractor?
60.
▲
by
ResearchCode
3y ago
Or as they call it outside of the agile industrial complex, a to-do list. Nobody is stopping you from keeping one.
More ›