4 ms·

> If they’re really so confident on the LLM’s effectiveness, why not just keep it voluntary, why force it on people? The results will be there in the outcome of

by rdtsc 1y ago

> If they’re really so confident on the LLM’s effectiveness, why not just keep it voluntary, why force it on people? The results will be there in the outcome of the shipped product for all to see. It’s a bit like returning to the office. If it’s such an obvious no-brainer performance booster with improved communication and collaboration, they wouldn’t have to force people to do it. Teams would chomp at the bit to do it to boost their own performance.

BiteCode_dev 1y ago
That's assuming this is most people's objective when they are at work. And even if it was, that's also assuming this benefit would be superior to the benefit of remote work for the individual.
vineyardmike 1y ago
I don't want to wade into the actual effectiveness of RTO nor LLMs at boosting productivity, but if you buy into the claims made by advocates, it seems pretty obvious that the "in office boosts communication" claim is only true if your coworker (the other side of the conversation) is in office. Not everyone has the same priorities, so you'd have to mandate compliance to see the benefits. Similarly, many people don't like learning new tools, and don't like changing their behavior. Especially if it's something they enjoy vs something good for the business. It's 2025 and people will have adamantly used vim for 25 years; some people aren't likely to change what they're comfortable with. Regardless of what is good for productivity (which vim may or may not be), developers are picky about their tools, and its hard to convince people to try new things. I think the assumption that people will choose to boost their own productivity is questionable, especially in the face of their own comfort or enjoyment, and if "the business" must wait for them to explore and discover it on their own time, they risk forgoing profits associated with that employee's work.
- gherkinnn 1y ago
  I don't see how using vim is in any way bad for business, what a terrible example. And I don't even use it myself. Your argument also hinges on "business" knowing what is good for productivity, which they generally don't. Admittedly, neither do many programmers, else we'd have a lot less k8s.
  - vidarh 1y ago
    Indeed, I detest vim but I think mentioning it detracted from the argument by showing why developers tend to not trust it when others try to dictate what is "good for the business" based on their own views rather than objective metrics.
- danielrothmann 1y ago
  You've got a point on RTO. Because it's a group behaviour, if you believe it will have positive effects, mandating it could be a way of jumpstarting the group dynamic. With LLMs, I'm not so sure. Seems more like an individual activity to me. Are some people resistant to new tools, sure. But a good tool does tend to diffuse naturally. I think LLMs are diffusing naturally too, but maybe not as fast as the AI-boosters would like. The mistake these managers are making is assuming it's a good tool for work that they're not qualified to assess.
  - rdtsc 1y ago
    > You've got a point on RTO. Because it's a group behaviour, if you believe it will have positive effects, mandating it could be a way of jumpstarting the group dynamic. Fair point! It was just the first recent example of "it's obviously better but we'll force you to do it" I could think of. In case of RTO I think it should have been left to individual small team. If one is so clearly it would have been clear in a few years time which teams work better and which didn't and how it depended on them being in the same office.
- rdtsc 1y ago
  > Not everyone has the same priorities, so you'd have to mandate compliance to see the benefits. I think it could have been left to individual small teams or smaller units. After some time it would have been obvious that teams who went to the office delivered better results. If the benefits are really that obvious as they are usually touted, it shouldn't take long (a year or two). > It's 2025 and people will have adamantly used vim for 25 years But if they are productive and delivers results they can still use vim. I can see controlling the APIs used, the programming languages allowed etc. But if they are productive with vim, let them use vim. > I think the assumption that people will choose to boost their own productivity is questionable, especially in the face of their own comfort or enjoyment, That's fair. I guess it depends on the types of people. I had in mind motivated people who would be glad to be more productive and deliver results quicker. Yeah, if people are not like that and are trying to do the least amount of work and just coast then it's a bigger issue. Office or not office, AI or no AI probably won't shift the needle by much.
forgotusername6 1y ago
There are psychological barriers to using a tool that diminishes the work you previously thought was complex.
- bagacrap 1y ago
  And there are delusions of grandeur that non engineers get from vibe coding.
Gigachad 1y ago
I’m lazy. I’d rather work from home even if the office is more productive because it’s easier for me to not have to go to the office. If the AI tools actually worked how they are marketed I’d use them because that’s less work for me to have to do. But they don’t.
- bagacrap 1y ago
  But presumably if you could get 4x done in the office you'd only have to work one week for every month you worked from home. Seems worth it even if lazy? I agree with the second paragraph. I seek out ways to be more productive, whether you call that work ethic or laziness. I wish AI or heck any tool could 100x me. But somehow AI is the first tool that's ever been forced on me by an MBA
brushfoot 1y ago
> Teams would chomp at the bit to do it to boost their own performance. This assumes teams care more about performance than comfort and convenience. Many teams care about both. And which one wins out can vary. It can even change over time. The question is how to incentivize what, and what methods are the most effective at doing that for a particular team at a given point in time.
- rdtsc 1y ago
  > And which one wins out can vary. It can even change over time. Well good reason to let each team pick. Teams that care about performance will eventually do a lot better. The claim with these mandate is that the benefits are "obvious". So if it's very obvious they should be visible fairly quickly.

by rdtsc 1y ago

BiteCode_dev 1y ago

That's assuming this is most people's objective when they are at work. And even if it was, that's also assuming this benefit would be superior to the benefit of remote work for the individual.

vineyardmike 1y ago

I don't want to wade into the actual effectiveness of RTO nor LLMs at boosting productivity, but if you buy into the claims made by advocates, it seems pretty obvious that the "in office boosts communication" claim is only true if your coworker (the other side of the conversation) is in office. Not everyone has the same priorities, so you'd have to mandate compliance to see the benefits. Similarly, many people don't like learning new tools, and don't like changing their behavior. Especially if it's something they enjoy vs something good for the business. It's 2025 and people will have adamantly used vim for 25 years; some people aren't likely to change what they're comfortable with. Regardless of what is good for productivity (which vim may or may not be), developers are picky about their tools, and its hard to convince people to try new things. I think the assumption that people will choose to boost their own productivity is questionable, especially in the face of their own comfort or enjoyment, and if "the business" must wait for them to explore and discover it on their own time, they risk forgoing profits associated with that employee's work.

gherkinnn 1y ago
I don't see how using vim is in any way bad for business, what a terrible example. And I don't even use it myself. Your argument also hinges on "business" knowing what is good for productivity, which they generally don't. Admittedly, neither do many programmers, else we'd have a lot less k8s.
- vidarh 1y ago
  Indeed, I detest vim but I think mentioning it detracted from the argument by showing why developers tend to not trust it when others try to dictate what is "good for the business" based on their own views rather than objective metrics.
danielrothmann 1y ago
You've got a point on RTO. Because it's a group behaviour, if you believe it will have positive effects, mandating it could be a way of jumpstarting the group dynamic. With LLMs, I'm not so sure. Seems more like an individual activity to me. Are some people resistant to new tools, sure. But a good tool does tend to diffuse naturally. I think LLMs are diffusing naturally too, but maybe not as fast as the AI-boosters would like. The mistake these managers are making is assuming it's a good tool for work that they're not qualified to assess.
- rdtsc 1y ago
  > You've got a point on RTO. Because it's a group behaviour, if you believe it will have positive effects, mandating it could be a way of jumpstarting the group dynamic. Fair point! It was just the first recent example of "it's obviously better but we'll force you to do it" I could think of. In case of RTO I think it should have been left to individual small team. If one is so clearly it would have been clear in a few years time which teams work better and which didn't and how it depended on them being in the same office.
rdtsc 1y ago
> Not everyone has the same priorities, so you'd have to mandate compliance to see the benefits. I think it could have been left to individual small teams or smaller units. After some time it would have been obvious that teams who went to the office delivered better results. If the benefits are really that obvious as they are usually touted, it shouldn't take long (a year or two). > It's 2025 and people will have adamantly used vim for 25 years But if they are productive and delivers results they can still use vim. I can see controlling the APIs used, the programming languages allowed etc. But if they are productive with vim, let them use vim. > I think the assumption that people will choose to boost their own productivity is questionable, especially in the face of their own comfort or enjoyment, That's fair. I guess it depends on the types of people. I had in mind motivated people who would be glad to be more productive and deliver results quicker. Yeah, if people are not like that and are trying to do the least amount of work and just coast then it's a bigger issue. Office or not office, AI or no AI probably won't shift the needle by much.

forgotusername6 1y ago

There are psychological barriers to using a tool that diminishes the work you previously thought was complex.

bagacrap 1y ago
And there are delusions of grandeur that non engineers get from vibe coding.

Gigachad 1y ago

I’m lazy. I’d rather work from home even if the office is more productive because it’s easier for me to not have to go to the office. If the AI tools actually worked how they are marketed I’d use them because that’s less work for me to have to do. But they don’t.

bagacrap 1y ago
But presumably if you could get 4x done in the office you'd only have to work one week for every month you worked from home. Seems worth it even if lazy? I agree with the second paragraph. I seek out ways to be more productive, whether you call that work ethic or laziness. I wish AI or heck any tool could 100x me. But somehow AI is the first tool that's ever been forced on me by an MBA

brushfoot 1y ago

> Teams would chomp at the bit to do it to boost their own performance. This assumes teams care more about performance than comfort and convenience. Many teams care about both. And which one wins out can vary. It can even change over time. The question is how to incentivize what, and what methods are the most effective at doing that for a particular team at a given point in time.

rdtsc 1y ago
> And which one wins out can vary. It can even change over time. Well good reason to let each team pick. Teams that care about performance will eventually do a lot better. The claim with these mandate is that the benefits are "obvious". So if it's very obvious they should be visible fairly quickly.