With cutting edge stuff you cannot say statements like "I tried X and found that it is not good at Y".
You could not get X to be good at Y, but it's not impossible someone else can.
497 karma · joined April 16, 2020
You could not get X to be good at Y, but it's not impossible someone else can.
Imagine a senior IC staffed with 4 juniors, and they spend 2 hours with each every day. Then the junior is left with 6 hours to think through what they were taught/told. This is very similar to LLM development except instead of context switching 3 times each day, the senior can skip over the 6 hours of independent time the junior required to absorb the changes. But it still takes the same amount of time to deliver the 4 separate projects.
I find the existence of LLM development deeply troubling for a long list of reasons. But refuting the claim that an LLM is similar in many ways to a junior dev is unsubstantiated
>It also is completely useless to perform many tasks that a Junior developer can.
And there are many things one junior could be helpful with that a different junior would be useless at.
With that as a metric, 1 Senior + 4 juniors cannot build the company with the scope you are describing.
A 50-eng company might have 1 CTO, 5 staff, 15 Seniors, and 29 juniors. So the proposition is you could cut the company in ~half but would still require the most-expensive aspects of running a company.
2) Yes, which is a separate issue. 1) Nearly all IC's should have the option to at least work remotely on some days, 2) All management should allow for their reports to take the day off if the work is complete. The fact that I get away with working a 4 days a week is a loop hole not a solution.
3) Realities of paywall articles. I'm not paying for the top ~20 news sites individual $5/month subscription. Nothing wrong with disclaiming I haven't read it and providing my thoughts on the subject. I'm highly doubtful the article contains anything new anyways, just rehashing the same half dozen arguments that have been around for 20 years, plus 1 or 2 new ones sourced from 2020. Probably includes a link to some recent study about some group trying it out and finding some success, which the author extrapolate's to the rest of the corporate world, and probably commentary from some CEO. You've read it, let me know how far off I am.
What is it you think happens on Saturday? Do you think those baristas are working 7 days/week?
That said, I'm not sure why this is always so popularly discussed in Eng forums. The reality is, our work is inherently async, and 90% of IC's could work 1 days/week if we magically squished all our work in that time frame. Many in practice get away with a 4 day work week. My company has a no-meeting Friday and I'm full remote, if I've finished my work I just quietly take it off.
So the reality is this: a 4 day work week for IC's means either asking them to work less or forcing them into a schedule they already had access to but opted to not take.
Why would a business be incentivized to have their staff work less or force them into an uncomfortable schedule that doens't adhere as nicely to the realities of parenthood or other adult commitments.
This is static data that only needs to be updated once or twice a year. Why not just make this an npm or py module
You wrote me original message with less words
I joined google in 2022 (and have since left), even as a newcomer it was obvious that not only was the golden era over, but the era after the golden era was winding down too. The atmosphere wasn't "The reckless innovation is over but lets make the product as great as we can" - it was "Just don't break anything and squeeze out 1 or 2% improvements where possible"
They would never do that.
Industry has a good track record of this.
It seems to break for ranges including 0 though
100 / -1~1 = -3550~3500
I think the most correct answer here is -inf~inf
A Makrov chain is just the path taken through the course of a markov process. The terms 'chain' and 'process' are sometimes conflated in this context, but this is the most common distinction I've seen. As such you can run a markov process for some number of steps N times, and then ask how many generated chains contain the property you are interested in. The process is memoryless but the chain is the result of the process and therefor contains memory.
I agree 'Random walk' is a superset of 'a markov process', but IMO when someone says Random walk - they normally make assumptions that qualify it as a markov chain. Therefor it's useful as a teaching to just call it a random walk.
Way out there, on your own. Fewer creature comforts, more getting stuff done on your own. No paying the connivence fee associated with cities, having everything nearby but paying double the rent.
Yes this sounds like an obscene edge case but it's part of the deal
I mean it's not like we're studying the resume and looking for inconstancies. Generally I just picked one specific line, one specific claim, and drilled into it. I generally want to give you a chance to do the whole "STAR" thing, but about some specific technical you claim to be one of your biggest wins.
I never have (or have done to me) called any of the references. That feels like a step too invasive for a generic senior engineer gig. Maybe for management I would care more, but I don't interview for or apply to those.
Re: 'resume prank', I generally consider a fancy college to be a detractor. It mostly shows your parents drilled you in highschool, and not much else. It's basically a moot point.
This isn't an innovation. It's just un-optimized bubble sort. Back in college when I was learning bubble sort, I implemented this exact sort as a type of "MVP". And then I added the other flags and stuff they reference to try to polish the turd.
I mean I'm glad someone proved it works, but I thought they proved this works like 40 years ago.
Augmented Chess: Normal chess with conventional ELO ranking system but you get additional penalties & rewards based on common bad practices, and recovering from bad positions.
> The first person that breaks from the book line loses 10% of the start time (unless book was not an option / the line was exhausted)
> Missing a forced checkmate forces you to wait 5% of the start time before your next move
> Achieving any principled position good grants some time (Passed pawn, connected rooks, rook / queen / bishop battery, etc)
> Doing any principled bad position loses some time (Knight on the rim, blocked bishop, king past the first rank in early game or mid game
----------------------
Continued Position: Chess but you continue a position from a high-level chess championship. There are a couple value-adds here
> Provide lower level players with a way to start a middle or end game positions after a highly skilled player followed all the correct principles. My theory is this will reinforce why those principles exists, how they can benefit you, etc.
> Provide high level players a way to be forced into positions out of their comfort zone / their preferred styles
> Provide differently-skilled players to continue play from unique positions with the desired amount of odds. So a GM might play a 1000 ELO player but starting from a position with -9 evaluation, etc.
If you have any ideas, comments, or feedback LMK.
Think Edison, more than Tesla.
Many decades later we say more accurately, he was a bad ass scientist who either sexually harassed or straight up raped most of his female mentees and was generally kinda racist (I mean, so was everyone back then. Still tho) and a general asshole.
I mean I don't really think there is any point in declaring anyone the best scientist ever. But he's firmly in whatever the top tier is when only considering scientific contributions.