The Incentives Always Win

Hello again my friend,

This is Liight Work Weekly, where I pull together a few stories from the week that I think actually matter. Not always because they are the biggest stories, but because they say something useful about where business, technology, culture, and work are heading.

This week, the theme is incentives.

Substack is moving toward sponsorships. The Enhanced Games tried to prove what enhancement could do. Cursor is turning AI coding from a developer tool into one of the most important business products in the world.

Different stories, same lesson.

People can say whatever they want at the beginning. Eventually, the incentive shows up.

If you understand incentives, you can usually see where a company, market, or person is heading before they say it out loud.

Let’s lock in.

Substack (our biggest competitor at beehiiv) is now expanding its sponsorship push. According to Axios, the company hired Dan Robbins as its first head of brand sponsorships and is building out a broader native sponsorship program with brands like Uber, Balenciaga, T-Mobile, Polymarket, and others. Business Insider also reported that Substack has been testing a sponsorship marketplace with select writers.

On the surface, this is a product update.

Underneath, it is an incentive story.

Substack built a huge part of its identity around subscriptions and being different from the ad-driven internet. That positioning worked. It was clean, simple, and emotionally powerful. Writers own the relationship. Readers pay directly. No algorithm. No attention casino. No ad model breaking everyone’s brain.

But creator businesses are messy. Not every publication can become a subscription business. Not every audience wants to pay. Not every writer wants to lock their best work behind a paywall. And if advertisers want access to trusted audiences, and creators want more ways to make money, someone is going to build the bridge.

My bias is obvious here. I work at beehiiv. But that is also why this story matters to me. beehiiv has been very clear about the fact that newsletters need more than one monetization path. Paid subscriptions are great, but sponsorships, boosts, referrals, ad networks, and marketplaces are not random side quests. They are infrastructure.

This is not really about Substack being hypocritical. That is too easy, and honestly not that interesting.

A platform can start with a belief. Then the market shows up with a stronger incentive. Creators want to earn more. Brands want distribution. Larger publishers want flexibility. The platform can either support it, or watch someone else do it.

The Enhanced Games were one of those ideas that sounded ridiculous until you thought about it for more than five seconds.

If athletes are already breaking records under the rules, what happens when elite athletes are allowed to openly use performance-enhancing drugs under medical supervision? The pitch was not subtle. Let athletes use testosterone, peptides, growth factors, anabolic agents, metabolic modulators, stimulants, and all the things that usually live behind scandal and suspicion, then see what human performance looks like without the pretending.

The incentive was obvious: attention, spectacle, money, and the promise that the human body still had another level to unlock.

The actual results were more complicated.

Kristian Gkolomeev swam the men’s 50m freestyle in 20.81 seconds, beating the listed world record time of 20.88, though the result will not be officially recognized because of the event’s rules and the use of banned performance enhancers. But across the rest of the event, the superhuman breakthrough did not really arrive. The Guardian covered the controversy around the timing, while SFGATE framed the broader event as underwhelming, with only one major record moment after all the hype.

That is the interesting part.

Enhancement clearly matters. I do not think the lesson is “performance drugs do nothing.” That would be silly. But it also does not magically turn every athlete into a world-record machine. You still need the right athlete, the right event, the right training, the right genetics, the right timing, the right equipment, and probably a thousand other things nobody can fully control.

The incentive was stronger than the outcome.

But I also would not write the whole thing off. The Enhanced Games may have underwhelmed as a sports spectacle, but the broader enhancement economy is not going away. Normal people are already moving in this direction through TRT, peptides, recovery protocols, sleep tracking, supplements, wearables, cold plunges, red light therapy, and whatever else gets packaged as optimization.

Doping may not guarantee world records.

But the desire to upgrade the body is becoming very normal.

And once that desire becomes normal, businesses will form around it.

If you do not code, it would be easy to skip the Cursor story. Best if you don’t.

Cursor started as an AI coding tool, basically an editor that lets developers work with AI directly inside the place they write software. Very techy.

But the story has moved far beyond that. SpaceX is reportedly buying Cursor’s parent company Anysphere for $60 billion, which Axios described as one of the largest acquisitions of a venture-backed startup ever. Business Insider also reported that Cursor crossed more than $1 billion in annualized revenue in under a year.

That is a stupidly fast curve and an absurd amount of money.

But the more important story is not the valuation. It is the incentive.

Software is the bottleneck for almost every ambitious idea. A founder wants to launch a new product, but engineering is busy. A growth team wants an internal tool, but it sits in a backlog. A company wants to automate a workflow, but the thing still needs to be scoped, built, reviewed, tested, and shipped. Even when the idea is obvious, the code is often where it slows down.

So the incentive is simple: make software move faster.

That is why AI coding tools matter outside of engineering. If work that used to require a specialist sitting down for hours can be handed to agents running in parallel, checked by humans, and moved through a company faster, the entire company changes. Roadmaps change. Hiring plans change. Internal tools change. The speed of experiments changes.

Given that Elon already has xAI, it’s extremely likely we see a new player competing at the frontier, with Claude and OpenAI (both of which are drowning in controversy anyway).

A few more places where the incentive is showing up:

The through-line this week is that reality keeps even the most ambitious companies grounded.

Thanks, as always, for reading.

Darwin