Sam Altman calls remote work 'experiment' one of tech industry's worst mistakes
fortune.com
fortune.com
The repercussion of that mistake is remote work.
I don't think he's wrong on principle - remote work is less dynamic and impactful than physical teams. But it's the "...and it has to be in San Francisco" part that reminds me of how out-of-touch these guys are. Are in-person meetings better? Yes. Are they worth commuting 3 hours every day through a city that hates you? No.
Hat tip to you, Legitster. Louder for the expensive seats upfront.
But what we get is a 1 hour commute each way, to have to find a place to sit in some sort of dynamic desk arrangement, in an open office setting that is noisy at the very least. And those meetings that are so much better in person? Good luck booking a meeting room. Where is the focus and increased productivity?
Even if this CEO's opinion was precisely the one that he expressed, it makes no sense at all to express it. Why would he potentially damage his company's standing by making unnecessarily inflammatory statements? What is the benefit for OpenAI from this?
To give him credit, he was saying something a bit more nuanced it just got chopped up by the headline.
> “I think definitely one of the tech industry’s worst mistakes in a long time was that everybody could go full remote forever, and startups didn’t need to be together in person and, you know, there was going to be no loss of creativity,” he told attendees. “I would say that the experiment on that is over, and the technology is not yet good enough that people can be full remote forever, particularly on startups.”
I'm biased, but I don't really agree with him. To me, an experiment collects data methodically and comes to a conclusion based solely on the outcome. I suspect he's just stating his own biases as though they were facts—unless there's some rigorous study I've not seen.
My bias is not only that I work better outside of an office, but that I was involved in a fully-remote startup a few years ago that was the most creative and productive environment I've ever worked in, so I have trouble believing those qualities are related to whether a company is distributed or not, and tend to think failures to achieve them at a high level in a remote environment is usually due to other factors.
Just like lack of data for the harmful effects of smoking in the 1940s should not have allowed people to blindly say "There's no proof it's bad for me, therefore it's not bad for me."
Who knows? We could have landed on the moon even earlier with remote work (assuming the tech was there of course) but the overwhelming evidence is that being in the same room has worked good enough in the past. (replace moon with James Web Telescope for a more time appropriate analogy).
I don't have hard data to support that one is better than the other. I can only draw on my past experiences, history, and first principles.
Of course, remote work can work and has worked. Just look at open source. But if I were to start a company that had to do really difficult hardware engineering, I would probably choose not to be remote. If I were to start a company that had to do really difficult software engineering, I would think about it. Why do I place such a high importance on quick communication for a hardware startup but discount that importance for a software startup?
Anyways, like many things in life there likely is not a true better option. It depends on many things. I will add that one argument I absolutely hate is when people complain about commutes. That is a separate problem that American cities have been reaping and one that we are actively sowing
Edit: Just to address the implication head on... The claim you are making is that it's so obvious everyone but shills will admit it.
But this is itself not true. It's a divisive topic among every demographic who harbors an opinion on the matter.
The jury is definitely still out and the story in the end is unlikely to be monolithic, but different along several dimensions.
The jury is literally not still out in any way shape or form.
Smoking causes cancer. that is harmful. period. The only controversy that is left is whether political power should be exercised to protect people from themselves. Which in the U.S. at least tends not to be a tenable direction.
Unless you're specifically commenting on the "unless you have data" part. Which ironically, tobacco and other chemical companies usually have reams of, but keep hidden, and fight the discovery and unsealing of tooth and nail.
Furthermore, no one wants to fund the side of the science that slits the goose that lays the golden egg's throat.
So more correctly stated is "Everything is hot air. Data doesn't matter as long as it never sees the light of day. When it does, have more money than the other guy to control the narrative."
We need to get back to building communities, making eye contact, seeing people, engaging in random conversations, getting to know your neighbors. And this includes working in person (albeit not necessarily every day).
If the goal is to go back to building communities, then I would be delighted to agree whenever OpenAI leases a building in downtown Tulsa.
Spare me the absurdly saccharine sentimentality.
To take the generally good idea of interaction and building your community, and extend it to cover being herded by the self-serving desires of megacorps seeking in-person office drones is grotesque.
I'd happily support people interacting more and forming new relationships, but in a personal, voluntary sense that doesn't involve being crammed into a giant box with others that you didn't pick, hectored by HR scum, and made to work with people who aren't your natural friends on productivity theater or simply things you could have also just done from home, where you choose who you spend time with.
Can you imagine someone stating the implication with a straight face?
"The way to build communities is to force everyone into an office building for 8 hours a day"