Middle management is really hard. Most people do not fit into neat categories of needing micromanagement or benefiting from completely hands-off management. Middle managers need enough domain knowledge to provide detailed suggestions, but also need to distance themselves enough from individual contribution that they can focus on scaling up their teammates. What should I be doing to help newer middle managers succeed in their new roles?

A calm day and a good night’s rest grants a very different perspective on things. Always wait until the morning before sending out a negative message. Try to listen more aggressively with an open mind with the goal of discovering the truth.

I feel like how I hold everybody together or fail to do so will define my leadership during this first chapter of our existence as a research team. Is there a winning move? How long should I give myself to turn things around before I throw in the towel? I have no idea and I’m not sure where to turn for advice.

Is there a way to do paid internships for experienced hires who are currently out of the job market? This seems like it would provide much more mutual signal compared to trying to come up with interview questions. Would this be considered a good or a bad interview process?

What is the right way to motivate researchers? On one hand, we want to keep the pressure up so that people feel a sense of urgency. On the other hand, we want to avoid burnout, and we also want people to drift a bit so that they can find new areas away from their local areas of expertise. Trading these off is difficult. One idea is to have is to have each researcher publish a research agenda and refresh it quarterly with a senior researcher, who can help load-balance topics across the team and may have some intuition on which paths are likely to be easier or more fruitful than others. As long as the research agenda is progressing on a quarterly basis towards the mutually agreed-upon goal at a reasonable pace, the research program is considered to be successful.

Can people learn creativity and attention to detail, or are those more or less fixed personality traits? Does being generally well-read cause somebody to be a better researcher, or is it merely correlated via showing one’s general interest in the world?

Of course, a single day doesn’t matter. Of course, it’s about long term growth. Of course, there are outliers and noise and you can only control so much. But none of this changes the fact that it hurts when things don’t go well, even for a couple of days. I have to toughen my skin a bit and find ways to immerse myself in projects instead of focusing on blips that I can’t control.

Doing it all over the second time around, I thought the initial changes would be chunkier. And yet the process has been the same as before. Slow, methodical, minor bumps in performance here and there. Wherever we look, we find a small improvement, but there are never any huge gains. Maybe this is just how the research process goes, whether it’s the first, second, or nth time.

Is there any wisdom in business management books? I’ve always written them off as pointless, but maybe I should give them a try. Wouldn’t it be insane to assume that, unlike every other skill that people must learn and practice mindfully, management is something you can either do or not do, and that it also happens to be something that I can do?

How is it even possible for us to discover new things? We compete against a faceless mass of researchers around the world, working around the clock on the same problems that we stare at fruitlessly day after day. It seems vain, almost contemptuous, for us to believe that tomorrow will be different; that tomorrow will bring a spark of inspiration that we will be able to nurture into something of significance, against all odds and against all of our prior experience. And yet, I must smile and remain the eternal optimist.

Does “fake it till you make it” also apply to creatives or is it reserved for blending in at desk jobs until you become Director of Product Strategy or something

A day of rest. I spent most of today eating and roaming in the city. I collected my thoughts again about the upcoming week, but those thoughts barely seem to have changed from last week. One day I’ll go back through my notes to see if they are really as consistent as they feel in the moment.

I’m excited for this week though. I feel like we are finally saturating our time with short-term projects that should add value immediately. It took a bit too long to get to this point. One takeaway is that we need better tools to evaluate isolated changes instead of relying on massive end-to-end tests. This lets us kill failed experiments faster so that we can refocus our time.

I might also try managing a bit more forcefully. Empirically, letting researchers dictate their own research agendas does not seem to be producing good results. Who is to say that providing forceful direction is any better, but it might be worth trying, at least for this stage of our lifecycle when we are still trying to get off of the ground.

Today was unproductive. I started the day listing out some goals for the week, but then quickly got distracted by life. And then when I finally wrapped up personal stuff for the day, I had to go back and fix some stuff with taxes. And now I’m finally done but it’s late and I’m tired. But if not me, then who? I will pick a random topic and try to dig in a bit for a half hour before bed.

The pain is knowing that you have to trick yourself into believing that there is still hope. Yes, you thought that your current approach was promising, more promising than all of the others you have tried. Yes, every test along the way showed results that were leaps and bounds better than the predecessor. But alas, one final test, and the goal remains just out of reach. And so you are back to brainstorming at the chalkboard. Not to worry. This time around you’re more experienced, more patient; you’ll cook up something even stronger.

Today, I was forced away from my desk by a doctor’s appointment. After the appointment, I glanced at my phone and saw no messages. The cold wind blasted my face. It felt like sandpaper rubbing my skin raw. I stopped by a bookstore and impulsively bought the first staff recommendation that my eyes settled on, aware that I was overtly judging the book by its cover. I took the book outside and sat on a small bench on the sidewalk, where I devoured a chocolate croissant. Still no messages. I ducked into the subway to return to my office. Today was going to be a good day.

I have only been doing this for 15 years. I am by no means an expert. I literally have no idea what I’m doing. Please stop looking at me like I have relevant experience.

Research continues. We haven’t made any improvements to the underlying model for some time, but we are working out some bugs in our metrics. The hope is that the improved metrics will narrow down some new areas to explore. It is just as likely (or maybe even more likely) that the metrics say nothing new or meaningful, in which case we are back to square one. The paths in front of us are shrouded in darkness. My experience, which has taken me thus far, has no bearing on these new possibilities. What should I do?

More frustration today. I’ve been laying awake at night trying to go through mental checklists of all of the different details we might be getting wrong in our feature definitions. The worst part is that it’s possible nothing is wrong in the model itself. It could be that there’s a mistake in the model evaluation code, or really anywhere in the pipeline up to us finally reviewing results. Or maybe it’s all correct and our current feature set is bad. But does that mean we should spend more time re-checking things, or should we spend more time trying to come up with new things?