A Repository of Cool Stuff For People Who Are Somewhat Like Us

the futility of giving universally good advice

It’s coming up on seven years since I first started writing an advice column for the Sunday newspapers.1 The main thing I have learned is that my job is impossible, because there is no such thing as universally good advice.

Is buying a house a ‘good’ idea? Well, I don’t know. Do you want a house? That seems kind of important. Does it give you more degrees of freedom, or fewer? What else would you have done with the payments? What interest rate can you get? How secure is your income?

It’s hard to squeeze all the necessary caveats into a 600 word newspaper column, and attempting to do so rarely makes for scintillating reading. The choices are to steamroll over nuance in pursuit of a pithy read—get those clicks coming in, the job is to make people feel something, it doesn’t matter whether it’s anger or curiosity or amusement or tribal loyalty— ‘Why All Homeowners Are Suckers’—or to equivocate endlessly and say nothing at all… ‘On the one hand…On the other hand…On the gripping hand…’ Scrupulosity is the mortal enemy of snappy opinion-writing, which is why I have this safe space where column inches are infinite.

So it’s partly the constraints of the format, but that’s not the main thing. The main thing is that people are not identical, and stubbornly refuse to share my exact views and preferences. Most advice ultimately boils down to ‘works for me, and for people like me’. To the extent that you are not like me, you will benefit from different advice; maybe even the exact opposite advice.

Slate Star Codex has a good bit about reversing any sage words of wisdom you hear:

[W]hen a young person is looking for job advice, I worry that all the artsy creative people whose heads are already way too high in the skies will be reading books by artsy creative people who urge them to follow their dreams, and so be even less mindful of the importance of a secure future. And all the hard-headed down-to-earth people will naturally gravitate toward reading Have A Very Secure Future By Going Into Business by Warren Buffett, and maybe never get reminded of the importance of following dreams.

Same problem for anyone reading my blog. If you’re a narcissist who believes the universe exists as some kind of magical lamp to fulfil your wishes, you will surely benefit from reading my post on the fetishization of ‘positive thinking’. But I don’t think I have many readers in this category. And maybe even the opposite—some of us need a dose of that Law of Attraction-style happy horseshit: indulge in some mild magical thinking, slip some rose-tinted glasses over the harshness of reality, stop acting persecuted because people have values other than truth-seeking.

If different people need to hear literally opposite advice, we have a problem. The implication is not that we have a 50/50 shot of getting it right. The odds are much worse than that, because of the sheer variety of preferences and circumstances.

‘You have to try Keto!’ your workmate says, stabbing meat-like objects out of a tupperware and huffing acetone fumes in your face. Is there a 50/50 chance this is a good recommendation? No, of course not: keto is wildly inappropriate for the vast majority of people, and about the bottom of the list of ‘things you should randomly try’, narrowly beating out incest and country dancing. Most advice you hear that is not personally tailored to you is terrible.

Even if it happens to be good advice, you still have to nail the timing. Stable personhood is an illusion; you are more like a succession of selves who all happen to share the same name and hang out in the same slowly-decaying meatsack. Which of the myriad versions of you needs to hear that advice?

There’s a whole genre of supposedly life-changing books which have left me cold—not because they’re bad, but because you have to read the right words, in the right place, at the right time. One person’s quake book is another person’s did-not-complete. If it seems banal now, it might have blown the socks off an earlier version; if it’s mystifying or impenetrable, it might yet appeal to a later version.

No-one is ever forcing you to act on a recommendation, this is just another data point to factor into your decision, and more information is always better. Right?

…no. Search costs are a thing that exists. Bad information is worse than no information, so we have a responsibility to find and boost genuine signals, rather than spray out more noise. As Tyler Cowen says, giving someone your favourite book imposes an obligation on them: now they have to read it, or skim enough to convincingly pretend they read it, or at least feel guilty about not having read it.

Format, context, preference, timing: unless all these factors happen to line up, specific recommendations are not much better than random monkey noises.

I say all of this as a prelude to the fact that I’m about to…lay a bunch of recommendations on you!

Deep Dish Recommends

There’s a tab at the top of the site, Recommended, for my favourite tools, services, apps, etc. I’ve just finished overhauling the list and culling most of my previous recommendations, which is what prompted me to write this post.

But it’s not all bad news: I also added a bunch of new suggestions and placeholders for things in the pipeline, to the point where I ended up with more than I started with.

I think recommendations are still genuinely useful in at least three situations:

1. (Almost) universally applicable

If you want to invest in passive index funds, signing up with Vanguard is objectively a great choice. Are there fringe cases where it’s not ideal? Yes. Maybe you could do some tax loss harvesting with a robo-advisor, or whatever. But this is a very good recommendation for most people, and it makes no difference whether they’re old, young, thin, fat, Republican, Democrat, etc. Same goes for the small handful of lifestyle interventions that have a ton of research behind them: they really do represent (almost) universally good advice.

The downside is that this category is unlikely to be revelatory: to the kind of people who regularly read finance blogs, my recommending Vanguard is like saying ‘hey, did you know exercise is good for you?’. But I still think there’s plenty of low-hanging fruit here, not least because the boring advice is regularly passed over for the One Weird Trick shiny stuff.

2. Thoroughly vetted

Some people asked what camera I use, so I obliged them by adding it to the list of my favourite tools. But…I’m not a photographer, and I haven’t tested a bunch of different cameras, so what am I really saying here? “Here’s this camera, it works for me.” People who know about these things confirm it is a good camera, but my chiming in is pretty much just noise. So I’ve removed this kind of thing from the list.

By contrast, I do have something to offer in my areas of interest—travel gear, health and fitness stuff, finances. I’ve done a bunch of experiments over the years, have a decent handle on the research behind it, and know what works and what doesn’t.

Even when I’m coming in blind, I often spend hours researching best practice, making spreadsheets to compare product features, etc. The signal is not super reliable compared to an expert, but at least it saves you from repeating all of my grunt work: ‘I spent a lot of time doing a sort of meta-review of other reviews, and this is what I landed on, for whatever that’s worth.’

3. Like-minded

Some recommendations really do boil down to: ‘well, it works for me, and people who are somewhat like me’. And… that’s fine! In fact, it might even be ideal.

This doesn’t work in mainstream media, because it’s impossible to appeal to everyone at once: comments sections are always full of people who disagree, and are not shy about telling you how much your idea sucks. But the hit rate is much better in smaller groups of people who have self-selected, e.g. your friends, hobby groups, blogs, niche media.

You and I must have something in common, or you wouldn’t be here. Maybe not a whole lot, but…something. I’ll occasionally recommend something that falls flat, or write posts that you find boring or irrelevant. But the signal-to-noise ratio doesn’t have to be perfect; it just has to be better than the alternative. If this is no longer the case, that’s why God gave us the [x] button.

Trying to appeal to everyone at once is the death knell for writing, art, business, dating, etc. And it goes both ways: if you put bland, generic things into the world, that’s exactly what you get back.

I am greedy and I want more coming back to me.

For example: about half of the best books I read each year are suggested by readers. Sometimes they’re duds. But that’s fine! Keep them coming. Recommendations from readers and friends and bloggers is still a much better filter than aggregate ratings from the public at large.2

Unlike other static pages on the site, I’m going to open up the comments section on the Recommended page. If you have a suggestion for something you love, or can fill one of the holes I’ve identified, please post it up. I would love to crowd-source recommendations, keep the page updated over time, and build a repository of Cool Stuff for People Who Are Somewhat Like Us.


  1. On hiatus for the coronavirus; possibly reinstated at some future point.
  2. As a wise man said, people like Coldplay and voted for the Nazis. You can’t trust people.