4 ms·
OpenAI CEO Sam Altman calls the remote work 'experiment' a mistake
- jalapenos 3y agoWell the most important thing is that the cat's out of the bag. So the self-serving rules-for-thee flying around the world executive hypocrites can only complain impotently that their serfs aren't attached to their allotments.
- mytailorisrich 3y agoRemote work requires management to adapt and it also requires good management. My experience with technical teams is that management is usually poor to start with and, in addition, it didn't adapt to remote work.
- ilaksh 3y agoI think if you're OpenAI, at this moment, then you could recruit top talent even if you said they had to work in Antarctica. Not sure what that proves. They are an exception, not the rule. I think that we will see deployment of more and more software for different levels of virtual workspace presence. I think it's fair to require some type of minimal live overlap one or more times a week. Whether that is just an hour being available in a chat room or something like a 3d desk in VR/mixed reality or video chat every day, etc. will be up for debate. I think that for office work it seems unlikely the majority will really go back to full-time literally physically in an office. But I think you should consider a full spectrum of virtual attendance rather than just the physical attendance. It does seem like you have to be pretty firm with monitoring work output and whatever live attendance or you will get many people trying to take advantage and just not do work or short-shift a job and even communication. That just requires better/different management that actually can understand the work product and process to judge rather than relying on literally seeing someone in a space to guess whether they are working.
- jalapenos 3y agoAs "current hottest thing", he could recruit them even if their office was to specifically be his personal bathroom, work and sleep. So he's really just talking his own book.
- raverbashing 3y agoSam is speedrunning the grifter persona it seems I think some of his statements (either current or future) will age rather poorly