Welcome back.

Have you thought about subscribing? It's free.

Overconfidence and AI

Human beings are often more effective when we’re a bit self-effacing. “I think,” “Perhaps,” or “I might be missing something, but…” are fine ways to give our assertions a chance to be considered.

The solar-powered LED calculator we used in school did no such thing. 6 x 7 is 42, no ifs, ands or buts.

Part of the magic of Google search was that it was not only cocky, it was often correct. The combination of its confidence and its utility made it feel like a miracle.

Of course, Google was never completely correct. It rarely found exactly the right page every time. That was left to us. But the aura of omnipotence persisted–in fact, when Google failed, we were supposed to blame evil black-hat SEO hackers, not an imperfect algorithm and a greedy monopolist.

And now, ChatGPT shows up with fully articulated assertions about anything we ask it.

I’m not surprised that one of the biggest criticisms we’re hearing, even from insightful pundits, is that it is too confident. That it announces without qualification that biryani is part of a traditional South Indian tiffin, but it’s not.

Would it make a difference if every single response began, “I’m just a beta of a program that doesn’t actually understand anything, but human brains jump to the conclusion that I do, so take this with a grain of salt…”

In fact, that’s our job.

When a simple, convenient bit of data shows up on your computer screen, take it with a grain of salt.

Not all email is spam.

Not all offers are scams.

And not all GPT3 responses are incorrect.

But it can’t hurt to insert your own preface before you accept it as true.

Overconfidence isn’t the AI’s problem. There are lots of cultural and economic shifts that it will cause. Our gullibility is one of the things we ought to keep in mind.

Bitterness is consistent

It will never let you down.

Bitterness is never-ending, impenetrable and refuses to negotiate. If you give it a chance, it will persist.

It lacks nuance or surprise. It’s simply a wall you can lean against, whenever you choose.

Consistency is all it has to offer, actually.

Amazon Smile gets a frown

I’m pretty sure how the first meetings went almost a decade ago:

“Well, we’re paying our affiliates 5% for referrals. If we pay charities a tenth of that and call it a donation, it’ll be great PR and we’ll also make a profit on every sale because we won’t need to pay a full commission…”

Amazon didn’t invent the online affiliate concept, but the company certainly turned it into a significant engine for growth. If a consumer bought a $100 item that someone in their program linked to, they would pay the referrer about $5. That meant they had little chance of actually turning a profit on that sale, but it led to millions of new customers, people who came back again and again. It also usually turned one sale into a few…

The end result is that Amazon has paid many billions of dollars in affiliate fees, and their affiliates (like Wirecutter, CoolTools, and formerly, Squidoo) sometimes built entire businesses around the simple idea that recommending a product and sending someone to the biggest online retailer to buy it could lead to income.

Smile turned more than a million non-profit charities into affiliates at a bargain rate. Now, that $100 purchase turned into forty cents sent to the cause of your choice.

I used to send you, my esteemed readers, to Smile links, but when I realized how little each donation was, I switched to geni.us, and so our donations went up 10x.

To date, Amazon Smile has sent charities more than $500,000,000. That’s not a ‘donation’ of course, it’s simply an allocation of marketing spend. And yet, half a billion dollars makes them a very significant donor in aggregate, one of the biggest corporate donors in the world.

So why cancel the program?

I have a few guesses:

  1. The new management at Amazon is aggressively streamlining programs and expenses to enable focus and increased profitability. Programs that involved a lot of people and time got harder to justify.
  2. They’ve always done a bad job of explaining the program, as evidenced by the fact that many people (including me) were surprised that the aggregate donation number was so big.
  3. Most charities aren’t very good affiliates. My guess is that few of the million recipients did much at all to persuade their donors to buy from the smile link that donated back to them.
  4. And perhaps most of all, as Amazon continues to dominate online retail, the affiliate program is probably less important strategically, and cutting costs on that front may seem like a simple way to increase short-term profit.

Western industrial culture has had a hard time thinking about charity all the way back to Community Chest and the United Way. If the ‘game’ is to maximize profit, then sending money to good causes seems to undermine that. Some people argue that companies have a corporate social responsibility, that the compact they sign with the community is to not simply take as much profit as possible but to invest for the long-term well-being of the places they work and the people they work with. Or, as may be the case here, it’s simply a marketing tool.


Not that long ago, you could make a living as a typist.

Technology keeps changing the world.

Now, you’re more likely to find a job doing something that seems a lot less mechanized. But that too will be programmatic soon enough.

PS here’s an important new book about perfectionism.

There are no stupid mistakes

There are mistakes. These are moments when reality teaches us something.

And there’s stupid. This is what happens when we refuse to learn from our mistakes.

“Don’t be stupid” is a fine mantra. It’s particularly apt when talking about cultural forces, political agendas and our thoughtless impulses.

It’s also useful to remember to “make productive and generous mistakes.” And to learn from them.

Jumping to conclusions

Maybe one piece of information is insufficient to get from where you were to where you just ended up.

When we gradually walk our way to conclusions, we’re a lot more likely to find something useful.

Leaping is best reserved for generous acts.

Generous and selfish

We’re often reminded that the best way to get what we want is to help other people get what they want.

That opening the door for others makes it more likely that others will open the door for us.

What makes it odd is the implication that if we don’t get, it’s hardly worth giving.

But what if we separated the two and simply chose generosity because we could?

It turns out that happy people are more likely to be generous. (Which implies that generous people are more likely to be happy). Not because they get something measurable in return, but simply because abundance is a choice. And making choices celebrates our agency and potential.

What an opportunity to bring justice and care and dignity to the people around us.

Happy birthday.

Your secret recipe

Perhaps it’s a proprietary manufacturing technique, or the fact that you add a little chickpea flour to your dosa batter.

Professionals often develop secret approaches and recipes that they use in their work.

It doesn’t matter. Even famous ones.

If it really matters to your competition, they’re going to figure it out.

And if they figure it out, you’ll succeed based on all the things that aren’t a secret–your care, effort and empathy. Time spent securing the secret in a vault is time wasted.

An event or a journey?

They’re easy to confuse.

An event happens at date certain, then it’s over, nothing more to be done.

A journey might include an event, but it’s bigger than that, and ongoing.

A wedding is an event, a marriage is a journey.

The week a book is published is an event, while the creation, publication and lifespan of the ideas in the book are a journey.

The focus and energy we lavish on events can easily distract us from the journeys we care about.

The platform and the curator

Who has their hand on the dial?

Talk with someone who works at Apple, Amazon, Google, Linkedin, Facebook, etc, and they’ll be happy to give you tips on how to work the platform to your advantage. How to get a bit more attention for your podcast or your website or your photos…

Unstated in this helpful posture is an unseen bias: They have a platform mindset.

This is the opposite of the thinking at a record label or a book publisher or a newspaper. They understand that their most important job is curation–choosing what goes on the front page.

Of course, platforms have long been curators, but they embraced the role instead of denying it. Radio station program directors decided what would be in heavy rotation and bookstore owners figured out what to put in the window or by the cash register. These platform leaders understood that their decision to promote something instead of everything was a key part of their job.

The platform mindset is sort of helpless. The algorithm is in charge, they aren’t. The data decides, they don’t.

In the short run, this bias feels helpful in a lot of ways. It eliminates gatekeeper errors. Since everything has a chance, being alert as a gatekeeper feels less important. When Decca famously turned down the Beatles, it was a mistake that cost them for decades. A platform executive doesn’t have to worry about this, since they can carry everyone and let the market sort itself out.

But there are two problems with this chosen, learned helplessness:

The first is that it’s simply not true. The algorithm doesn’t write itself. The rise in hate speech on platforms like Twitter is possible because the algorithm rewards it. The vapid recipes that people build on websites are there because Google’s algorithm rewards them. And yes, the noxious additional fees that airlines charge are there because the travel websites rank flights with hidden charges higher in the results than those that are honest about what it really costs.

When platforms grow in scale, they often add hardworking, well-meaning people to engage the public, sort of a buffer between creators and the algorithm, but they’re instructed that the algorithm itself is sacrosanct and off limits. We sell everything, we don’t know how to sell any (particular) thing.

Go to some meetings with Apple’s podcast team (the ones who apparently can make or break a podcast by promoting it) and you’ll soon realize that Apple isn’t really in the business of helping its many users find podcasts that will elevate, inspire and educate them. Instead, they’re simply feeding the platform.

Netflix, in its best moments, succeeds because they break the platform paradigm and shift into curation.

The second is that this platform-first agnostic non-curation ultimately leads to the demise of the platform. The aphorism is: Enough A/B testing will turn any website into a porn site. That’s because the short-term waves of data-driven, algorithmic feedback loops inevitably make platforms banal, then trashy. This is happening to Amazon–their Amazon Go stores in New York are dull. The search result spam on their site is worse. Inevitably, the people they most want to serve get frustrated, bored or bummed out and go somewhere else.

What to do about it? Well, if you’re a creator, it helps to realize that you’re probably not going to be treated in a special way by any platform that has a platform mindset. There isn’t a shortcut, there are simply lots of dreary steps and then maybe some luck.

And if you’re part of a platform that has scale (or hope to build one), this is the perfect moment to learn from the curation that came before. When we talk about the folks that built the parts of our culture that we’re proud of, we almost never talk about the platforms. We talk about people who had the guts and the taste and the energy to help others discover things that made a difference, all while winnowing out the cruft and the junk.

We shouldn’t be here to feed the platform. The platform needs to be here for us.