Worthwhile Content: May 2026
Some worthwhile reads, watches, and listens from the last month
World of Work:
In May, I wrote a slightly darker piece arguing that Companies Like AI Because It Allows Them to Work The Way They’re Actually Designed To Work, showcased an excerpt from my forthcoming book with Tips to Make Sure Your Mentoring Conversation is Actually Helpful, and responded to reader Colin’s question about his company taking all his best people and distributing them elsewhere.
Some other worthwhile content I digested in May includes:
I’m saddened by the death of the Sloan Management Review. Rita McGrath gives it a nice obit and talks about how HBR is really all that’s left to translate academic work into practitioner insights. Which is a bummer, because many of management’s pressing questions already have useful answers that the general public doesn’t know about.
New Cornell research confirms what most of us have suspected: people who love corporate jargon are worse thinkers and decision-makers. The twist is a bummer, though, which is that these people are actually more satisfied at work. Far from suffering, they're thriving, and potentially elevating the kinds of leaders who use the stuff.
This interactive visualization maps the meta-analytic relationships between cognitive abilities and personality traits based on Stanek & Ones' landmark research. If you've ever wanted to see how, say, conscientiousness relates to fluid intelligence, or how openness maps onto various cognitive domains — all in one place, all corrected for unreliability — this is the map. Genuinely cool for anyone who spends time thinking about individual differences.
If you work for yourself, or want to, check out this Substack on 20 things you should know if you work for yourself.
Check out my latest monthly appearance on Truth, Lies & Work! Maybe one day the Brits will learn to spell organizational correctly.
AI & Work
Greenhouse surveyed nearly 3,000 job seekers across five countries on AI interviews. Many have had one, most weren’t told beforehand, and AI didn’t convince candidates of decreased bias compared to humans. As Greenhouse's own CEO put it, layering AI on top of a broken process doesn't fix it.
Ethan Mollick highlighted a new paper that looked at how AI compares to human writing.
Mary Kate Stimmler showcased that organizations which track and report AI value are far more (20x!) likely to achieve high returns. Shockingly, if you measure something transparently, you tend to get more value out of it!
Owen McGrann’s piece entitled “The Dead Economy Theory” is extremely thought-provoking, if not demoralizing. And is definitely worth a read.
General Interest
New York Magazine's profile of Tim McCormack is a terrific, albeit tragic story of a sports gambling compulsion and the betting scandal involving NBA G-League player Jontay Porter. Porter’s co-conspirators bet on him having a lousy statistical game, so he left the game “injured” ensuring that everyone would win their bet. (paywall)
Speaking of gambling, the WSJ dug into the data behind Polymarket and Kalshi and found that 67% of Polymarket profits go to 0.1% of accounts. The winners are algorithmic trading firms running tens of thousands of trades a day against casual bettors who mostly just bet on what they hope will happen. I am growingly convinced that these bet-on-anything websites are hugely problematic and in need of tighter regulation, to say the least. (paywall)
We’ve solved the cross-over between eastern and western medicine?! This is wild and completely makes sense: Apparently, we’ve discovered that the body has a third circulatory system (in addition to the cardiovascular and lymphatic systems). Acupuncture is based on this (probably). We’ve long known acupuncture works and this provides the best evidence as to why. And it’s really cool.
If you have not used Pangram Labs for AI detection you should. I use it regularly (you’d be surprised just how many of your favorite follows are using 100% AI text!). But it has its flaws: I have tested my own writing while working on my book and found it to say that some parts were likely written by AI (narrator: they weren’t). In my own trial, the AI-flagging is likely owing to the highly-structured, bullet-heavy sections of the book, which makes some sense as that style is not as common in typical human writing. As The Atlantic points out, a lot is riding on Pangram.
Scientific criticism is a cornerstone of good science, and this is a clean example of how to do it. Daniel Lakens writes a detailed response to Amy Cuddy’s insistence that her Power Posing research holds up. I don’t know what to make of Amy Cuddy (never met her) and her clinging to the research findings so tightly, despite the fact that the evidence is scant and one of her original co-authors has backed away from their initial conclusions. I spoke about said study at my recent Lecture on Tap.
In one of the most face-palm moments this year, a hot new book about how AI is corrupting truth was found to contain more than half a dozen fake or AI-generated quotes. The author, Steven Rosenbaum, acknowledged using ChatGPT and Claude in the research process and owned the mistakes. It doesn’t strike me as malintent, just laziness. If AI generates a citation or a quote and you’re writing a book, just go have it verified. It takes a minute to find sources.
Similarly, Harvard Business School is suing their ex-professor, who is alleged to have fabricated data in her work to make it more impressive. The story is so compelling because of the focus of her research: honesty.
I really enjoyed this episode of the Galaxy Brain podcast where they discussed what it’s like to shift away from smartphones. Speaking of which, I’ll be shutting down social media for these first 10 days of June. Join me?
A former WilmerHale lawyer makes the case that Big Law is broken. He argues that our current system cannot protect the democratic ideals the practice of law is intended to protect. 2025 was a record year for the industry; meanwhile, firms are caving to poorly-written legal action from the government.
This essay about a father who looked back at the times he texted his kids that they were being “too loud” and now just wishes he could hear their voices after they’ve left the nest was quite lovely. You-don’t-know-what-you’ve-got-till-it’s-gone is an old trope, but not a tired one.
If you’re a millennial, read this piece on our “credentialed class problem”. A bit painful just how much this hit me square in the nose! Dang. Dead on.
Victor Wembanyama is headed to the NBA finals (GO SPURS!) and is also quickly becoming one of my absolutely favorite athletes, if not outright humans. His maturity and thoughtfulness at such a young age would be impressive for anyone — for someone who has grown up in a microscope and was expected to be the next Kareem Abdul-Jabbar, it is nothing short of astounding. I love this guy.
Musings
I’ve noticed a subtle distinction. I don’t mind reading AI content on places like LinkedIn when it’s supposed to be informational. A post written by AI that simply shares a useful insight doesn’t bother me (though I’d still like it to be labeled). What irks me is the AI content that’s supposed to be conversational. That’s when it feels inauthentic. Don’t fake being human.
Here’s a hot take (I can’t remember if I shared this before): If you capture video in public that shows strangers, you should be required to get their consent before posting. I could see a carve out for certain locations (e.g. the way one can shoot video in Times Square without prior authorization), or exceptions for posting something that could be deemed a “demonstrable public good”, such as showing something illegal, unethical, or of deemed public value. But when it’s just for fun, I’d be down for consent or forced facial-blurring.
That’s it for this edition - please reach out if I can be at all helpful.
Be compassionate and deliberate.