Crackpot time 3: speculations will turn out well?

In 2017 I wrote two posts about my about my experiences with attempting to do physics outside of academia, which I called Crackpot Time 1 and Crackpot Time 2. At the time I was trying to reconnect to a more expansive, free-ranging energy that I had lost during the hyperfocus on technical details required for Ph.D. work. I was enjoying the ‘crackpot’ label as a kind of tongue-in-cheek pointer to the style of thinking I was trying to cultivate. I wanted to directly attack any topic that looked interesting, without fussing about whether the topic was ‘too ambitious’, or ‘too difficult’, or ‘not my field’. Small details like a total lack of relevant expertise didn’t matter.

I had a lot of this kind of energy in 2017, which was a very good year intellectually for me. I went to two deeply unusual and inspiring physics workshops that immediately raised my ambitions for what it would be possible for me to do in my spare time alongside a full time job. At the same time I was starting to take my side interest in mathematical intuition more seriously, and get oriented reading some phenomenology for the first time, so it was an intense time where I felt like the horizon was opening up fast in all directions. I started this blog and cranked out a bunch of short, unpolished but enthusiastic blog posts to try and make some sense of my thoughts.

I’ve been meaning to write another Crackpot Time update ever since, but just… never have. Partly that’s because I started a monthly newsletter practice in 2018 that took over some of the same role. But also it’s the standard inspiring workshop problem: the inspired feeling eventually wears off and then you then have to do the hard bit, which is doing the actual work. This is less immediately exciting and doesn’t autogenerate breathless updates about how amazing everything is, so they stopped appearing. I’ve finally decided to crank one out anyway, even if it’s effortful and uninspired.

At the beginning of 2020 I got this fortune cracker for Chinese New Year. Perfect fortune for a crackpot, right?

I’m now trying to evaluate whether speculations did in fact turn out well. It’s weirdly hard to decide. I’m normally at least somewhat confused by my progress – trying to do independent work in a complicated domain is slow and ambiguous at the best of times – but I think this is the most confused I’ve been in a long time. Long 2020 has obviously been enormously strange for everyone, and then on top of that I’m in a hard-to-interpret stuck phase. This is my best attempt to explain what I’ve been up to, and where I’m at now.

Focus and accountability

I’m not going to try and go over everything I’ve done since 2017, nobody cares including me, but I’ll do a few quick catch-up paragraphs to get me to the beginning of 2020. I had two good strategic ideas at the start of 2018. The first was to pick a very specific topic to focus on. My natural tendency is to dissipate my energies going partway down some interesting rabbithole before getting distracted by something else, and only end up with a very vague high-level understanding of anything. Useful for getting a sense of the territory, useless for making any sort of meaningful contribution to it.

To counteract that, I picked a single 8-page paper, A toy model for quantum mechanics by S. J. van Enk, as my focus for the whole year. I had some sense that this particular paper would be a good anchor for me, and that turned out to be correct. The core toy model is very concrete and easy to play around with, but touches on a number of ideas in quantum foundations that interest me – negative probabilities, the phase space formulation of quantum physics, the Spekkens toy model. There are also potential intriguing connections to my favourite recurring fascination, retrocausal interpretations of quantum physics. Having to stick close to the anchor paper meant I could explore aspects of these big topics without disappearing off into uselessly ungrounded speculation.

The second good idea was to use a monthly email newsletter as an accountability mechanism, inspired by this post. This wasn’t a Substack or anything, just a bog standard email that I sent out to a handful of people. I’d ramble a bit about what I’d done in the month, and that gave me a bit more incentive to stay on track. I stuck fairly closely to the area of this paper for the whole of 2018 and didn’t stray much further in 2019 either. This gave me far more focussed knowledge than I’d managed to pick up before working on my own.

At the beginning of the year I wrote the following:

My plan for 2018 is to go beyond just learning some physics in my spare time and to do ‘something novel’, interpreted broadly. ‘Novel’ in this case doesn’t have to mean original research (though that would definitely count) – I’m thinking of a wider conception of what counts as a novel contribution, in the style of Chris Olah and Shan Carter’s Research debt essay (I wrote some comments on it here).

I’ve never been too fussed about whether anything I do is original in the sense required for an academic physics paper, as a completely new technical contribution to the field. But my ambitions are higher than just passively making notes from a textbook. I want to follow my own curiosity trail through a subject, write down what I notice on the way, and highlight ideas and connections that currently aren’t available in digested blog post form. The sort of work that Olah and Carter call ‘research distillation’ in the essay linked above.

This took longer to spin up than I was initially hoping for, and I spent most of 2018 just learning background and writing notes. I finally got going in 2019 and had a few thoughts on negative probabilities from a somewhat novel angle, which produced a couple of posts and a mildly popular twitter thread. So that takes me up to 2020, and the fortune cookie.

Long 2020

In early 2020 I had a tedious hour+ two-bus commute to work and sometimes skimmed some interesting-looking papers on my phone. Otherwise I wasn’t getting much done, because my energy was sapped by the stupid commute. I decided to have a twitter break in February to claw back whatever time I could, which worked fairly well. Some time near the end I spent a Saturday holed up in a corner of Bath University library where I had an idea for a very basic toy model that was quite limited by itself but maybe extensible in some interesting way. I was excited to figure out what it could do and started fiddling around with that for the next week or two.

I got back on twitter on March 1 to discover everything had been replaced by coronavirus panic, which was a big shock to me because I had almost completely ignored it until then. So I started catching up on panic, and the toy model went out of my head for the next couple of months along with everything else that wasn’t covid. I no longer had the bus commute, but I also couldn’t think properly, so that didn’t help much.

After a couple of months my brain came back online at least partially, but the toy model was completely dropped. (I still haven’t managed to pay it any consistent attention, it’s a loose thread at the moment.) Instead I remembered the papers I’d been reading on the bus. I’d been learning about Abramsky and Hardy’s logical Bell inequality work, and I realised that I could use the tools from this to finish off a half-baked idea for a post on Bell’s theorem that I had, connecting a classic popular-science explanation to the version you’d find in a text book. The logical Bell inequality techniques made a natural bridge between the two, and over the summer I was able to use this idea to extend my scrappy notes into a full post that I was pretty happy with. I was finally managing the kind of distillation work I’d been thinking about at the start of 2018.

After that I was on a roll, and found a second use for the logical Bell techniques. In my 2019 posts on negative probabilities I used a very simple toy model created by Dan Piponi as an illustrative example. I picked it because it was simple, but I was also intrigued by its relation to quantum physics – it’s structurally similar to qubit phase space, but the specific numbers are different. In a sense it’s even further from classical physics, with the negative probability being more negative than anything allowed in quantum physics.

I’d noticed before that this was interestingly parallel to a much more well-known case of something being ‘worse than quantum physics’, the Popescu-Rohrlich box, but thought it was only a vague similarity. Once I had the logical Bell tools I realised that there was an exact numerical analogy. I couldn’t find this described anywhere else, so I started writing it up.

Unfortunately this took long enough that it took me into the long depressing UK lockdown winter. The news was a constant stream of miserable statistics from the new covid variant mixed in with increasingly batshit US election nonsense, the weather was dark and grey, and working from home was getting more and more tedious. I eventually managed to finish the ‘worse than quantum mechanics’ stuff and get it out as two blog posts, but that overstressed my limited ability to care about things and once I published the posts I lost interest. I made some very half-hearted attempts to find out more about whether this was actually novel, and when this wasn’t completely straightforward I just dropped it. That was some time around February and I still haven’t picked it up again.

So… now what?

I’m writing this up now because I suddenly have a lot of free time. I’ve just quit my job – last day was last Friday – and haven’t lined up another one. I’m planning at least a couple of months off before I start thinking seriously about getting a new job. So this would be the perfect time to pick this up again. I’m not too bothered if I can’t get my attention back round to physics, because I have other weird projects that I am still keen to work on, but it does seem like a shame to just drop all this stuff. I’m not going to push it though.

The thing I’m feeling most is the lack of social support. I’m not naturally plugged in to a community of people in quantum foundations who are thinking about similar topics, so it can be difficult to keep motivation. David MacIver has a great newsletter post on Maintaining Niche Interests, where he talks about struggling with the same problem:

“Nobody actually wants to know” is a bit unfair. It’s more like… there are people who are interested, but they are both less interested than I am in the subject, and also I don’t talk to them much. The people who I talk to on a regular basis are not interested, because this is mostly not their field.

I feel it even more keenly in comparison with some of my other interests that I talk about on this blog and newsletter and on Twitter, where I do have some sort of community. I can talk about some pretty niche topics – Derrida, Vygotsky, the Prussian education system – and get meaningful informed responses from other people. Book recommendations, suggestions for related areas to explore, that sort of thing. It’s not the same as being in a densely-networked in-person research group, but it goes a surprisingly long way.

The pandemic has definitely made it worse. I do normally get some sense of shared community from the physics society I’m in, which organises workshops and meetups (including the two really inspiring ones I went to in 2017). But it’s very much a community built around meeting in person, rather than around producing large quantities of English-language text on the public internet. We’ve tried a few online calls and talks, but it’s not the same.

Even without the pandemic, though, I struggle with this. I’m just not very good at collaborating when it comes to physics. A lot of this is rooted in defensiveness – I’m just weird for a physicist, kind of slow and mediocre technically and with an odd thinking style, highly focussed on examples and weak on abstraction. I go into any interaction worrying that I’m going to look stupid and expecting to not be able to get my point across, which makes it even harder to get my point across, which… you get the idea. It’s difficult. I think I could make good incremental progress on this in the same way I made progress on blogging, but getting the right supportive environment to start the feedback loop going is tricky. Physics culture is not known for providing what I want.

In the mean time I’m going to keep plugging on with other projects and not force anything. After all, it’s been a strange enough year that I should probably feel happy that I did anything at all. Hopefully my interest in physics will return soon and I can get a better sense of whether speculations have turned out well.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s