A friend of mine working as a deep learning engineer at one of the big tech companies is working her ass off due to fear of layoffs. She mentioned that their manager takes quiz! from them and gives them performance tests. The vibe is incredibly tense and except a few who are working on secret projects, the rest are left hanging in a looming lay-off round.
Every day she works till late at nights or awake 5am to continue working. She now thinks that working like that over the pas few months has reduced the chances of her being laid off but she's is completely fed-up and I think she's risking having a burnout.
Sounds like she has a bad or even toxic manager. A big problem research
orgs have is the 'publish or perish' mentality imported from academia. The business version of this is the "what is the last thing you did for me?" mentality which is toxic.
Managers need to be more Ted Lasso than Led Tasso.
6h30 to 8h a day (that's 37h a week for me). I like my job and I have fun but It's still to much for me. I'm looking into a 4 days week kind of contract.
I'm a cofounder of a Series A start-up, so work hours are hard to define. You are sort of always working or thinking about work and it is hard to put away.
But within the company, we've found lots of uses for LLMs to save time. One or two of the developers are fans of Github Copilot, but that is a relatively minor impact so far. The bigger impacts are around data cleaning and content. A lot of jobs that used to be "hire someone on Upwork for $20/hr to do a lot of boring clean-up work on thousands of records or write repetitive text about something" are now "write a 10 line python script with the OpenAI api to generate the result for $8 and then QA". There are also plenty of uses of boring old home-grown ML models and stuff like that to clean up data in cases where the problem is more specific and predictable.
So far AI just makes our best developers faster and our best product people make a larger impact more quickly and at a lower cost. It hasn't resulted in less internal hiring or anything like that. But it has resulted in less Upwork-style one-off contracting, especially around data cleaning and content clean-up.
2 times 8 hours a week, software made my life at work worse (at first it looked like a really good deal then the Government managed to convert my database software to database service with some extra responsibilities on me about uptime). The layoff wave have not touched me yet but since I'm a male Ukrainian, the danger of mobilization is hanging over me which will make my working hours 24 times 7 times 365 times life.
If considering what's called "Deep work" only. Then 3-4 hours max. Otherwise, whole day.
AI has definitely helped me save my time debugging issues. Otherwise, there has been no increase in pressure. Rather, I have stopped preparing hard for interviews.
doomscrolling is a competitive environment; HN is a decent start so you are already at a good starting point, although you might want to level-up to FB, or even TikTok for more advanced skills.
37 hrs/week, of which 3/day might be solid work, the rest is eye rolling and facepalming about things like salesforce, oracle and similar horror straight from hell.
Not affected by layoffs, working for a company that is a moat, that virtually prints money.
AI saves me time as a context-heavy stackoverflow at hand, with high signal to noise ratio. It won't replace me any time soon. Excited for AGI though, I could use intelligent colleagues.
No, I don't work more because of the layoff waves. As a recent father with scars from a close-to-burnout situation, I value work-life balance more than anything else.
So far, I'm not using AI much, mostly because I lack the routine to use it, but I'm not opposed to using it if/when it can help me.
I work 30 hour weeks at my day job, works out to 6 hours a day with change. Then I spend anywhere between 0 and 10 hours more working on my personal projects.
I'm battling a chronic illness at the same time as I'm running my own company with a couple of employees.
The positive side is that my health issues forces me to delegate and not get involved in day-to-day operations too much (which I previously did). Now I can easily work 1-2 hours a day at max and still keep the company humming along. Most work is on how to design the business long term for success and often not directly involved with clients.
AI helps immensely to quickly get things out of the way and only focus on essential things. Many mundane tasks can be done using AI now and it really is a game changer and makes it easier to take on more complex tasks as a business owner.
Probably 2-4 hours of work related stuff and thinking about work
Meant to be doing 1+ hours of learning and reading each day.
Trying to take weekends more seriously but that depends on $ for the week. Daily consistency is a big challenge for me. Some days I can just fall off a cliff. Others it will be 10-11 logged hours.
Copilot and Chat every day while programming. Little tasks and boiler plate are taken care of. When getting very deep into developing a system, the chat needs more and more context so becomes less useful.
I don’t know how many hours I work, but recently I saw that in the last week I was in front of a screen for almost 12 hours on average (per day)… And that’s with going out for lunch and sleeping. I wasn’t even working for a lot of those hours. It’s been a long cold winter here in the north. Currently I’m rapidly introducing some corrections, as I felt I was a wreck.
I have a 30 hours per week contract. Though I sometimes find it hard to actually work 30 hours and not much more (when I am trying to tackle an interesting problem).
I have a four hour work-week (30h), and work on average 6h a day, sometimes 6.5-7 if I have something I'm in the middle off. Certainly used ChatGPT to save some time, but hasn't affected my working hours.
Good point, I've seen top performers laid off because of (what I assume) comes down to a cost-benefit analysis. Top performers may get paid significantly more, and may present more of a challenge to management (or challenge management more).
My layoff survival strategy comes down to costing too much to get rid of. I do that by mastering and turning into the go-to for some critical part of the infrastructure the other developers shun. Most often that means the database, but it could mean the cloud setup or even details of the business domain. If no one else in the organization can take over my work without significant time and cost I will get scratched off the layoff spreadsheet. I survived a few layoffs that way, watching more senior people get walked out.
Of course when your employer starts laying off you should start planning your own exit, but avoiding the first wave or two will give you some breathing room.
Professor, so maybe not who you mean to ask. Four hours of deep work, that is, creative work. Maybe another eight of stuff. No AI yet, although I'm super interested.
AI only wasted my time. It cannot really solve anything beyond basic StackOverflow questions and it even fails on super simple questions too. So I wasted hours to discover the fact.
I work as long as I want in theory, but I try to work 8hr/day, which is expected from me, but of course at least 2hr/day is spent on meetings. Sometimes I don't "work" at all (I mean coding), usually I get around 2-4h to think and code.
I'm not worried about layoffs, while I usually don't believe in myself I feel I still have a strong skills that are needed on the market and the worst that can happen is a little smaller salary in another company.
That lines up with my AI experience, too. These days the entirety of my AI experience is sitting on huddles with juniors/intermediates who will type out half a line of code and then wait for Copilot ad infinitum. It can be a bit annoying--just write the code yourself!
I think these tools have some time saving applications for writing boilerplate, but I've never been able to use them to actually solve a real problem. ChatGPT feels like an improvement over Copilot but there's still a ways to go.
I'm interested to see how Copilot X [1] turns out. Maybe it will be more useful.
Trying out entrepreneurship for the first time, and looking at my coding activity tracker I put in around 7-9 hours of coding, with perhaps 1-2 hours on top for non coding activities, each day.
I usually take a day off each week with little to no work.
66 comments
[ 2.7 ms ] story [ 139 ms ] threadEvery day she works till late at nights or awake 5am to continue working. She now thinks that working like that over the pas few months has reduced the chances of her being laid off but she's is completely fed-up and I think she's risking having a burnout.
I always get suspicious of mass layoffs -- I always think they are not connected to performance but to networking and marketing of your work.
Managers need to be more Ted Lasso than Led Tasso.
But within the company, we've found lots of uses for LLMs to save time. One or two of the developers are fans of Github Copilot, but that is a relatively minor impact so far. The bigger impacts are around data cleaning and content. A lot of jobs that used to be "hire someone on Upwork for $20/hr to do a lot of boring clean-up work on thousands of records or write repetitive text about something" are now "write a 10 line python script with the OpenAI api to generate the result for $8 and then QA". There are also plenty of uses of boring old home-grown ML models and stuff like that to clean up data in cases where the problem is more specific and predictable.
So far AI just makes our best developers faster and our best product people make a larger impact more quickly and at a lower cost. It hasn't resulted in less internal hiring or anything like that. But it has resulted in less Upwork-style one-off contracting, especially around data cleaning and content clean-up.
If considering what's called "Deep work" only. Then 3-4 hours max. Otherwise, whole day.
AI has definitely helped me save my time debugging issues. Otherwise, there has been no increase in pressure. Rather, I have stopped preparing hard for interviews.
If I have other busywork I do as much as necessary, if not doomscrolling or read literature.
Most projects have bug trackers with tags for beginner friendly bugs.
37 hrs/week, of which 3/day might be solid work, the rest is eye rolling and facepalming about things like salesforce, oracle and similar horror straight from hell.
Not affected by layoffs, working for a company that is a moat, that virtually prints money.
AI saves me time as a context-heavy stackoverflow at hand, with high signal to noise ratio. It won't replace me any time soon. Excited for AGI though, I could use intelligent colleagues.
I work 30 hour weeks at my day job, works out to 6 hours a day with change. Then I spend anywhere between 0 and 10 hours more working on my personal projects.
The positive side is that my health issues forces me to delegate and not get involved in day-to-day operations too much (which I previously did). Now I can easily work 1-2 hours a day at max and still keep the company humming along. Most work is on how to design the business long term for success and often not directly involved with clients.
AI helps immensely to quickly get things out of the way and only focus on essential things. Many mundane tasks can be done using AI now and it really is a game changer and makes it easier to take on more complex tasks as a business owner.
3ish hours of buisness work a day.
Probably 2-4 hours of work related stuff and thinking about work
Meant to be doing 1+ hours of learning and reading each day.
Trying to take weekends more seriously but that depends on $ for the week. Daily consistency is a big challenge for me. Some days I can just fall off a cliff. Others it will be 10-11 logged hours.
Copilot and Chat every day while programming. Little tasks and boiler plate are taken care of. When getting very deep into developing a system, the chat needs more and more context so becomes less useful.
Some days, just nothing happens. Can be spooky quiet.
> Did the layoff wave push you to work more hours than expected?
No. It is unlikely to affect me.
> Does AI save you time already?
No. It's done the opposite - I spend a lot of time either fielding questions about how we can use it (in most cases we can't, yet).
> Did the layoff wave push you to work more hours than expected?
No. I am a top performer (not a self assessment, officially documented) so I figure I'll be the last to go.
False sense of security. Quite a few other things can take priority over being a top performer, when it comes to cost cutting.
My layoff survival strategy comes down to costing too much to get rid of. I do that by mastering and turning into the go-to for some critical part of the infrastructure the other developers shun. Most often that means the database, but it could mean the cloud setup or even details of the business domain. If no one else in the organization can take over my work without significant time and cost I will get scratched off the layoff spreadsheet. I survived a few layoffs that way, watching more senior people get walked out.
Of course when your employer starts laying off you should start planning your own exit, but avoiding the first wave or two will give you some breathing room.
I work as long as I want in theory, but I try to work 8hr/day, which is expected from me, but of course at least 2hr/day is spent on meetings. Sometimes I don't "work" at all (I mean coding), usually I get around 2-4h to think and code.
I'm not worried about layoffs, while I usually don't believe in myself I feel I still have a strong skills that are needed on the market and the worst that can happen is a little smaller salary in another company.
I think these tools have some time saving applications for writing boilerplate, but I've never been able to use them to actually solve a real problem. ChatGPT feels like an improvement over Copilot but there's still a ways to go.
I'm interested to see how Copilot X [1] turns out. Maybe it will be more useful.
[1] https://github.com/features/preview/copilot-x
I usually take a day off each week with little to no work.