Knowledge Work's Model T Moment

From 1900 to 2000, the productivity of the manual laborer in the US increased by a factor of 50. Thanks to this, we went from 40% of the population working in agriculture to 2% today, yet have more food available than ever.

Almost half of US workers now work in the "knowledge work" sector, where the rise of the internet has led to lots of changes in how we work in the last 30 years. With all these changes, are we seeing similar productivity improvements here? No, argues Cal Newport in his book "A World Without Email: Reimagining Work in an Age of Communication Overload".

According to Newport, thanks to the internet, new workplaces revolve around the "Hyperactive Hive Mind". This is a "workflow centered around ongoing conversation fueled by unstructured and unscheduled messages" on platforms like email and Slack. Newport argues that this working style has actually reduced productivity instead of boosting it, as we just spend more time on communication with no clear benefit to productivity.

Anyone who's worked in a modern company can relate. Keeping up with the flurry of Slack messages and emails feels like a full-time job, leaving little time to get actual work done. Do you feel like we are more productive today than 20 years ago? The data suggests no, and my personal experience says the same.

We still haven't found our "Model T" moment for knowledge work

Newport argues that despite all the "innovation" we've seen, knowledge worker productivity today is similar to manual labor productivity before the assembly line in 1900. In his view, we still haven't found our "Model T" moment and we are just scratching the surface of how to organize and use knowledge workers.

One example Newport brings up is a small startup which has a (WSJ described) "radical" policy of a 5 hour workday:

"They arrive at around eight each morning and leave at around one in the afternoon. During the day, social media is banned, meetings highly restricted, and email checks constrained. When they’re done with work, they’re actually done until the next morning—no late-night sessions at the keyboard, no surreptitious smartphone messaging during their kids’ sporting events—as professional efforts are restricted to time spent in the physical office."

Now I know many people (myself included) who think this is a really bad idea for most companies and especially startups. 5 hours is just not enough time to get things done, among other issues.

But Newport's more important here point here is not that this is the clearly the right model for all companies to move towards, but that there is a shocking lack of experimentation in how companies run.

It’s now widely accepted that continued industrial growth requires continual experimentation and reinvention of the processes that produce the stuff we sell... As Drucker reminds the reader, since 1900 the productivity of the manual laborer increased by a factor of fifty!...“On this achievement rest all of the economic and social gains of the 20th century.”

But when we turn our attention back to knowledge work, we find this same spirit of experimentation and reinvention lacking. This is what I meant when I wrote in the Times that the rarity of experiments like Lasse Rheingans’s five-hour workday was itself “radical.” Rheingans is thinking about his organization with a Henry Ford mindset, by which I mean he’s looking for bold new ways to deploy his capital to produce more value."

Is Newport right? Let's focus on tech companies, the clearest example of the knowledge work economy and a sector I know well. Is there really little "experimentation and reinvention" in how these "innovative" companies operate?

When I look at it, it's hard to conclude anything but "yes." Tech company culture has become pretty homogenized, with everyone having the same (generous) benefit policies, casual dress code, and open floor plan. They also have the same boilerplate values, all so agreeable that they become functionally meaningless (who can disagree with "We do the right thing"?). Google's "weird" culture, which maybe was innovative at one point, now seems mainstream, outdated, and cliche all at the same time.

Company hierarchies/organization also look pretty similar, regardless of how "flat" they profess to be. Companies hire in similar ways, both in the process (ex: call-> test -> final round) and in what they look for (Stanford engineer, ex-Stripe). If you are qualified for one name brand company, you tend to be qualified everywhere else.

And, as Newport emphasizes, almost all tech companies revolve around the exact same type of email/slack powered hive mind, have the same daily structure (8-10 hour days, Mon-Fri), and use the same tools, meetings, and processes.

Who is innovating in the knowledge worker space?

Obviously, each company is unique in its own way and there are some companies which seem to be truly innovating. Amazon is first to come to mind with its business workflows like the 6-page memo, PR FAQs, 2 pizza rule, and Day 1 culture (great overview here). Netflix treats its employees like a "team," not a "family," and awards generous severance to those that don't perform (more on this in another post).

But honestly these examples don't impress me that much. Consider this: if Amazon's culture is so different, why do you see a rather similar revolving door of people go between Amazon and other big tech companies? The work and culture is far from foreign to the average tech employee. I'd argue moving companies today is like moving cities in the US. You can move from NYC to Chicago and things will be slightly different, but you'll get the hang of things quickly and you can adjust. But what's the equivalent of moving countries? What companies are so different in how they operate that it feels like moving to Shanghai, not Chicago?

One example of a company culture that sounds truly foreign comes from outside of tech: Bridgewater. In case you aren't familiar, Bridgewater, led by founder/CEO Ray Dalio, has a culture of "radical transparency" which entails things like: employees constantly rating each other on 60 personal attributes, assigning employees "believability ratings" for how to weigh their vote in decisions, recording all meetings and sharing for the whole company to watch. In one stark example, co-CEO Eileen Murray was investigated as to whether she lied on a "trivial" matter (if she or her assistant wrote an email) and, well...

"Videos of the interrogations were edited and later rolled out in a serialized fashion, a Bridgewater version of a reality TV show, said the former employees who saw them. The videos also served as a case study as part of “homework,” a firm practice in which employees review and analyze recorded meetings and internal debates." (Source)

Now that is a weird culture! I can't imagine any company I've worked at doing anything close. Dalio compares it to first walking into a "nudist camp" and says it takes about 18 months to get adjusted to the culture.

To be clear, I'm not here to argue Bridgewater's culture is good (I have mixed views). It's clearly worked for them (arguably the most successful hedge fund ever), but there are many legitimate critics and Dalio himself argues it's not for everyone. But we should want 100 more companies like Bridgewater who are genuinely trying new methods of organizing knowledge workers. Only with experimentation and trying new things can we really find better ways of working together.

Remote work and AI to the rescue?

I've been largely negative on knowledge sector innovation, but I do think remote work is a noteworthy current exception and AI may change things drastically in the future. Having a remote-first company with no office is a genuinely new way of working together and could be the most important innovation of the next decade. Even before COVID, there were notable examples of successful companies using this setup like Gitlab, Zapier, and Toptal. And companies worth many multiples of these have decided to go remote-first since the pandemic, though many have been reversing. At least they are trying!

Remote work has led to more innovation in how we work beyond just physically not being in the same place. For example, Gitlab runs entirely on a written down in a handbook and meetings are rare or often optional. The culture is so transparent that anyone can read the handbook, including all of Gitlab's current and historical OKRs (!!!).

Ironically, the change many companies embraced was forced upon them against their will. I don't think most companies going remote-first seriously considered it pre-COVID, and yet they have radically changed the way they work in favor of something that (they think) is better as a result. This further supports the view that there are many hidden innovations that companies can experiment with for knowledge workers, but inertia (among other things) is holding them back. Whether you are bullish remote work or not, you should welcome this experimentation.

And then there's AI. AI has not had a big impact on how we do knowledge work yet, but I expect it will in the future. New ways of working will be possible with AI assisting us, likely in ways we haven't ever considered.

Will we think about company headcount in terms of humans and AI agents? Will we be serving up decisions for an AI to answer and execute on? Will there be companies with tens of millions of revenue and one employee? Will seamless translation mean every company becomes global? I think this will be the biggest opportunity for knowledge work innovation in the next decade.

Is culture a "solved problem"?

Is company culture just a "solved problem?" That was one suggestion a friend had when discussing these ideas. It being a "solved problem" would explain the homogeneity and the reason many companies have decided to copy the dominant culture instead of trying to create a new wacky one. Silicon Valley is replete with "best practices" on how to run a company, so why try to reinvent the wheel?

After all, how much evidence is there that innovating on culture actually leads to better outcomes? Bridgewater has done well, but so have 20 other hedge funds with traditional cultures. Amazon has also excelled, but there are many successful tech companies today that haven't really done anything but copy.

We may not be on the cusp of the new "Model T" for knowledge work, but I do think it's far too early to call this solved, especially with AI around the corner. While everyone may have copied Google and FB in the last decade, I expect there to be new companies that we emulate in the next 10 years that work very differently.

And one day, we'll get the Model T moment for knowledge work where we look back on the "Hyperactive Hive Mind" like we do pre-assembly line manufacturing and say "I can't believe we worked like that."