Thinking About #4 ... Repeating the Past, Behavior and Productivity
Sharing a number of things I've been thinking about lately. Getting them out of my head. This time is about the tradeoffs in repeating history, behavior-linked problems and the myths of productivity.
This post is part of my “Thinking About Series”. This is a semi-regular post where I get all the thoughts bouncing around my head onto digital paper. I’m averaging one every 7 to 10 days covering 3 themes. I don’t have a fixed publishing schedule or template.
When should we repeat history?
Crypto constantly reinvented or rediscovered things that finance had been doing for centuries. Sometimes it found new and better ways to do things. Often it found worse ways, heading down dead ends that traditional finance tried decades ago, with hilarious results. Often it hit on more or less the same solutions that traditional finance figured out, but with new names and new explanations.
From Matt Levine’s Crypto Story
… Feyerabend argues that the history of science is so complex that it cannot be reduced to a general methodology; asserting a general method will inevitably inhibit scientific progress, as any unifying and static method would enforce restrictive conditions on new theories…
From How to Accelerate Science - City Journal
I’ve been thinking about reinvention and recreating history/the past a lot recently. My working career has involved rediscovering things that people before me had done. Many ‘trends’ within the startup and tech space are just repackaging of previous trends, research or management styles. Companies setup innovation and incubation functions to separate the new stuff from the legacy structures yet those internal-startups end-up building the same structures. Companies struggle with knowledge management and retention. When your boss changes, it's normal to start all over again so they can be seen as “owning” the projects.
And within these massive structures, change slows to a crawl. New technologies take years to incorporate into the company way of working. Even positively impactful change has to go through so many rounds of approval, most people just give up and maintain the status quo. “It any broken. Don’t fix it.” New methodologies and approaches are rolled out regularly to combat this - Agile, LEAN, Disruptive Innovation… Yet in time those approaches face the same decline. Inevitable mistakes lead to risk management lead to bureaucracy which eventually leads back to a team doing well in-spite of the process.
Where is the balance? In what situations does it make sense to ignore the research, history and practices and reinvent things and when does it make sense to slow down and learn from what has happened before?
This is especially challenging because people get used to practices and form habits. Processes are developed based on mistakes and risk management. As a company or a team becomes larger, it becomes less risk tolerant (more to protect and more exposure when things go wrong). More stakeholders join teams with wider mandates and skin-in-the-game leading to ever more checks. Even governments become more process driven and can end up being very efficient at producing somewhat negative outcomes.
“Forget the past and we are doomed to repeat it,” as the saying goes. Yet remembering everything about the past requires a level of process, control and cost which isn't always ideal. Times change, technologies develop, people have different preferences yet if we continue to run systems to avoid past problems we will deliver negative outcomes. We need to find the places where repeating the past is valuable. Even forgetting that we have tried this before may be more valuable than keeping track of what happened and why.
Whenever faced with a problem, we can solve it by following proven historic processes which will take time but likely guarantee a reasonable, iterative solution. We can throw ourselves in and try to solve it fresh which could lead to failure or a fresh, new solution that no one has thought of before. Like I’ve talked about before, I wonder how much of this lives in areas beyond or normal perception. Our daily life is habitual and we don’t actively question “this didn’t work before, is this worth another try?” and even the value of willful ignorance: “who cares if its been tried before!”.
We follow process when we shouldn’t. We retain knowledge that is useless. We focus on the wrong problems because the right problems are hard, uncomfortable. The wrong problems are easy to talk about and never solved.
Behavior dependent problems we avoid
I recently joined a public policy student mentorship program. I was impressed by the enthusiasm of the students and their eagerness to solve society’s unsolvable problems. At the opening event, we discussed the inactivity of the richest people in the world to solve these impossibly complex problems.
“If only X would direct their billions to solving Y problem”… It is fun to take digs at people with more money than my entire ancestry and future generations will even conceive of. Later on, I got to thinking about why. With so much money, these people must have some perception of these problems and are choosing not to solve them. Why?
I wonder how much of it is due to the complexity of the problems and that the problems are highly dependent on human behavior. World hunger isn’t a problem that rockets and scientists and money can solve. It has a horrible mesh-up of political, social, cultural, individual, geopolitical influences and no easy solutions.
As the Bill and Melinda Gates Foundation continues to demonstrate, even endless money and the most intelligent people in the room struggle to eliminate the most addressable diseases. These are the problems that humble gods.
Billionaires aren’t dumb people and clearly achieved their positions through a cocktail of luck, timing and targeting solvable problems that make money. These titans of industry have selected problems where three things will accelerate a solution: money, risk-tolerance and expertise. These are problems distant from human behavior. You don’t need to change human behavior to get paid to launch rockets, or sell cars, or deliver same-day, or offer music.
The horrible irony of this exploration is that I can see myself, all of us, in a version of this same reality. How often have all of us sat down with our friends and tried to address a social issue in our neighborhood, family or company? Almost never. We avoid problems that are dependent on human behavior. Its too uncomfortable, random and risky. Best to focus on what we can impact through our efforts.
Seen another way, if we would spend that money trying to solve these behavior-dependent, complex problems, we wouldn’t have made that money to begin with. Those rich-f*cks are doing what they do best, what made them rich and what they enjoy - they are shoving capital, expertise, and embracing risk in an area where those things have a chance of paying off. Defuse, structural and behavioral problems don’t have that upside.
We all laughed at the extremism of Softbank’s Son funding of WeWork and other start ups that ended up failing or producing sub-par returns. To Son and all those who backed him and the thousands of funds, founders and startups like him - they are laughing all the way.
This is fun for them. This is what they do. Find a problem where money, expertise and risk-tolerance will produce a solution and do it. Ignore all the problems that have human-behavior tied up in them.
Solutions for social problems are seen as a side-effect of a successful business; saying things like: “my company will make a lot of money that I will spend and that will give people jobs, solving poverty for those who choose to work.”
Which leaves me thinking about what do we do with all the behavior problems that we are left with?
Productivity isn’t a value-case/use-case
A lot of people obsess about productivity - how it is measured, what it means or doesn’t mean for the economy and now - GenAI. Since ChatGPT, GenAI has been touted as a major boost for productivity and many studies have shown that use of GenAI tools helps with a wide range of tasks, increasing quality and productivity.
Even with this research and the reality that at least 50% of a given professional workforce is using GenAI tools daily (or lying that they aren’t), compelling use-cases are generally lacking beyond impacts on freelancers.
The implementation and use case gap in GenAI is because we still don't have good ways of quantifying productivity improvements in most knowledge-worker environments. Productivity doesn't stand alone by itself as a good value-case org wide. It is highly appealing on an individual level as the wide-spread adoption of GenAI demonstrates.
Corporate use-cases are still dominated by tangible benefits like reduced costs or increased revenues. Until GenAI applications mature to the point that they can show implementations that result in direct cost savings in excess of the implementation and running costs or some how demonstrate increased revenues, they will remain tools that are more important to individuals than to organizations. As I consider this, I think this remains a concern for a wide range of digital projects as a whole.
Fundamentally, I think that companies are missing an opportunity to experiment with completely different role, team and delivery design by rebuilding approaches with GenAI. That will happen eventually as it did with computers and other productivity tools, but for now we are stuck in a messy middle. Companies that are running ahead with GenAI tools and use-cases are doing it with the hope of a payoff or as part of an organizational leadership driven fear-of-missing-out.
This all is making me think about when that tipping point will be. It seems like it will be a mix of traction via really break-out compelling applications and it will be iterative and organic with most companies seeing incremental improvements in the tools they use. There will certainly be a few break-out companies shifting the dynamic in the market based on unique approaches that eventually become benchmarks that the rest of the market tries to emulate (badly) like what happened with things like the Spotify-Method.
Conclusions
I’ll continue to explore these and other topics in future “thinking about” posts. Do let me know your additional perspectives and we can continue to explore.