I like Mike Hall’s tips for interviewers in his notes about behavioral interviews. The crux of the behavioral interview style is this:

I’d rather know what you’ve done than what you think, and I have adjusted my style a little to help candidates. I try to explain my process up front. “I’m going to ask you about times you did things because I really want to get down to how you work and what your experience is.”

It is, however, very important to realize that adjacent or related experiences are completely valid and should be encouraged:

If you want to build a diverse, vibrant team, or if you’re not one of those disasters of a manager who doesn’t understand that you need people at several levels of experience on a well-rounded team, then you need to think of a behavioral style not as a way to narrowly insist on stories that describe the exact thing you need done. Instead, you need to think in terms of the competencies the thing requires, and think of examples to ask for that reflect those competencies, not an exact task.

This is a good tip for candidates as well—don’t talk in generalities. Be specific about the ways in which you have solved a particular challenge in the past, what worked well, and what you have learned/would do differently next time.

This isn’t entirely surprising but it’s a sad state of affairs, and it’s worth highlighting not just how, but also where LLMs are being trained:

Hundreds of thousands of hours of work goes into providing enough feedback to turn an LLM into a useful chatbot, and that means the large AI companies outsource the work to parts of the global south, where anglophonic knowledge workers are cheap to hire.

I know it’s too dismissive to call chatbots “fancy autocomplete” like many do, but we have to remember that this isn’t magic. The words the bots use come from somewhere. And in the case of “delve”…

I said “delve” was overused by ChatGPT compared to the internet at large. But there’s one part of the internet where “delve” is a much more common word: the African web. In Nigeria, “delve” is much more frequently used in business English than it is in England or the US. So the workers training their systems provided examples of input and output that used the same language, eventually ending up with an AI system that writes slightly like an African.

Mandy Brown nails it, once again:

One of the principles I come back to over and over is adrienne maree brown’s invitation to move at the speed of trust. That is, whenever attempting any effort with other people, prioritize building trust and respect for each other over and above any other goal. The trust forms the foundation from which the work can grow.

If you’re a fan of the Murderbot series (and if you haven’t read it, get on it!) you will absolutely love this recent keynote speech by author Martha Wells at the annual Jack Williamson Lecture at Eastern New Mexico University. She describes how Murderbot came to be, what it’s really about, and where the story sits within sci-fi and our world in general:

There are a lot of people who viewed All Systems Red as a cute robot story. Which was very weird to me, since I thought I was writing a story about slavery and personhood and bodily autonomy. But humans have always been really good at ignoring things we don’t want to pay attention to. Which is also a theme in the Murderbot series.

I won’t ruin the ending by quoting the final paragraph, I’ll just say that this is my favorite thing I’ve read in a long time, and you should savor every word.

I admit that I love post-apocalyptic books and movies. So I feel more than a little bit chastised by ‘End of the world vibes’: why culture can’t stop thinking about apocalypse:

“Such convictions in the mouths of safe, comfortable people playing at crisis, alienation, apocalypse and desperation, make me sick,” complains the protagonist of Saul Bellow’s 1964 novel Herzog. “We must get it out of our heads that this is a doomed time, that we are waiting for the end, and the rest of it … Things are grim enough without these shivery games … We love apocalypses too much.”

Read on for more choice quotes from olden times that make the case that “the only way to manage a dread of the future is to remember that the past was no picnic”.

Michał Poczwardowski shares a good reminder about how to set expectations well in our teams:

The biggest partner in crime for missed expectations is unclear communication, which means that the antidote is clear communication. Follow these steps to make sure that expectations are clear:

• Be realistic about the future. Overconfidence will build up expectations. If there are a lot of uncertainties, state clearly what is certain and what is uncertain.
• Point out what you don’t know. Give as much context as you can. If you leave too many unaddressed gaps, people will fill these gaps with their own projections of which you have no control over.

This is a good reminder about micro-management from Judd Antin. He says that maybe You’re Not Managing Enough (a big climbing analogy runs through the whole post):

As managers, we can be so afraid of micro-management that we risk moving into passive territory. We’re made to believe our main job is putting people in position to grow, and then going hands off to give them the space to do it. But that’s like encouraging a climber to take on a harder route, cheering them as they start while you check out TikTok instead of holding fast to the rope. To do their best, that climber needs an active belay from start to finish. It’s easy to try again when the rope caught you and you only fell a few feet. These are the most educational failures — it’s those big ones that you want to avoid.

There’s some practical advice in the post on the best ways to be more active and helpful in the right ways by providing clarity and making solid plans with your team.

I like this alternative approach to the old “disagree and commit” adage. That idea always struck me as a little passive aggressive (“sure, I’ll do this stupid thing you want me to do…”), whereas this seems like a more active, helpful approach:

It’s much healthier to “have concerns and commit.” Some decisions you can agree with, some you can disagree with, but most you should either just “have concerns about” or “be supportive of”. […] If you’re not sure of the answer but have concerns, you want to make sure that your feedback is deeply considered. You can tell your team that feedback was heard but ultimately the people with the most context made the call, which is how it should be.

It’s important to note that this type of culture is only possible if leaders agree to provide a lot of context on decisions (which not everyone wants to do):

As much as you have to be humble in your approach to engaging with decisions, healthy companies and leaders should provide you with enough information to be able to understand decisions in enough detail to have confidence in supporting the decision.

I found this essay on why we do things that are bad for us really interesting. First, I learned the word “akrasia”, which means “the state of mind in which someone acts against their better judgment through weakness of will.” Second, this is not exactly a new thing. From Romans 7:18–19:

For I have the desire to do what is good, but I cannot carry it out. For I do not do the good I want to do, but the evil I do not want to do—this I keep on doing.

Anyway, I thought this was helpful advice:

To achieve your goals, it can be more effective to put into place a defined plan that doesn’t let you reconsider. The psychologist Peter Gollwitzer called this an implementation intention: come up with a specific if/then statement that helps you achieve your goal. If it’s Tuesday, then I will go to yoga class; if I buy spinach, then I will make this smoothie for breakfast the next morning.