Channelling

I’m writing this five minutes after making my wife’s “coffee decoction” using the Bialetti Moka pot. I don’t like chicory coffee early in the morning, and I’m trying to not have coffee soon after I wake up, so I haven’t made mine yet.

While I was filling the coffee into the Moka Pot, I was thinking of the concept of channelling. Basically, if you try to pack the moka pot too tight with coffee powder, then the steam (that goes through the beans, thus extracting the caffeine) takes the easy way out – it tries to create a coffee-less channel to pass through, rather than do the hard work of extracting coffee as it passes through the layer of coffee.

I’m talking about steam here – water vapour, to be precise. It is as lifeless as it could get. It is the gaseous form of a colourless odourless shapeless liquid. Yet, it shows the seeming “intelligence” of taking the easy way out. Fundamentally this is just physics.

This is not an isolated case. Last week, at work, I was wondering why some algorithm was returning a “negative cost” (I’m using local search for that, and after a few iterations, I found that the algorithm is rapidly taking the cost – which is supposed to be strictly positive – into deep negative territory). Upon careful investigation (thankfully it didn’t take too long), it transpired that there was a penalty cost which increased non-linearly with some parameter. And the algo had “figured” that if this parameter went really high, the penalty cost would go negative (basically I hadn’t done a good job of defining the penalty well). And so would take this channel.

Again, this algorithm has none of the supposedly scary “AI” or “ML” in it. It is a good old rule-based system, where I’ve defined all the parameters and only the hard work of finding the optimal solution is left to the algo. And yet, it “channelled”.

Basically, you don’t need to have got a good reason for taking the easy way out now. It is not even human, or “animal” to do that – it is simply a physical fact. When there exists an easier path, you simply take that – whether you are an “AI” or an algorithm or just steam!

I’ll leave you with this algo that decided to recognise sheep by looking for meadows (this is rather old stuff).

Order of guests’ arrival

When I’m visiting someone’s house and they have an accessible bookshelf, one of the things I do is to go check out the books they have. There is no particular motivation, but it’s just become a habit. Sometimes it serves as conversation starters (or digressors). Sometimes it helps me understand them better. Most of the time it’s just entertaining.

So at a friend’s party last night, I found this book on Graph Theory. I just asked my hosts whose book it was, got the answer and put it back.

As many of you know, whenever we host a party, we use graph theory to prepare the guest list. My learning from last night’s party, though, is that you should not only use graph theory to decide WHO to invite, but also to adjust the times you tell people so that the party has the best outcome possible for most people.

With the full benefit of hindsight, the social network at last night’s party looked approximately like this. Rather, this is my interpretation of the social network based on my knowledge of people’s affiliation networks.

This is approximate, and I’ve collapsed each family to one dot. Basically it was one very large clique, and two or three other families (I told you this was approximate) who were largely only known to the hosts. We were one of the families that were not part of the large clique.

This was not the first such party I was attending, btw. I remember this other party from 2018 or so which was almost identical in terms of the social network – one very large clique, and then a handful of families only known to the hosts. In fact, as it happens, the large clique from the 2018 party and from yesterday’s party were from the same affiliation network, but that is only a coincidence.

Thinking about it, we ended up rather enjoying ourselves at last night’s party. I remember getting comfortable fairly quickly, and that mood carrying on through the evening. Conversations were mostly fun, and I found myself connecting adequately with most other guests. There was no need to get drunk. As we drove back peacefully in the night, my wife and daughter echoed my sentiments about the party – they had enjoyed themselves as well.

This was in marked contrast with the 2018 party with the largely similar social network structure (and dominant affiliation network). There we had found ourselves rather disconnected, unable to make conversation with anyone. Again, all three of us had felt similarly. So what was different yesterday compared to the 2018 party?

I think it had to do with the order of arrival. Yesterday, we were the second family to arrive at the party, and from a strict affiliation group perspective, the family that had preceded us at the party wasn’t part of the large clique affiliation network (though they knew most of the clique from beforehand). In that sense, we started the party on an equal footing – us, the hosts and this other family, with no subgroup dominating.

The conversation had already started flowing among the adults (the kids were in a separate room) when the next set of guests (some of them from the large clique arrived), and the assimilation was seamless. Soon everyone else arrived as well.

The point I’m trying to make here is that because the non-large-clique guests had arrived first, they had had a chance to settle into the party before the clique came in. This meant that they (non-clique) had had a chance to settle down without letting the party get too cliquey. That worked out brilliantly.

In contrast, in the 2018 party, we had ended up going rather late which meant that the clique was already in action, and a lot of the conversation had been clique-specific. This meant that we had struggled to fit in and never really settled, and just went through the motions and returned.

I’m reminded of another party WE had hosted back in 2012, where there was a large clique and a small clique. The small clique had arrived first, and by the theory in this post, should have assimilated well into the party. However, as the large clique came in, the small clique had sort of ended up withdrawing into itself, and I remember having had to make an effort to balance the conversation between all guests, and it not being particularly stress-free for me.

The difference there was that there were TWO cliques with me as cut-vertex.  Yesterday, if you took out the hosts (cut-vertex), you would largely have one large clique and a few isolated nodes. And the isolated nodes coming in first meant they assimilated both with one another and with the party overall, and the party went well!

And now that I’ve figured out this principle, I might break my head further at the next party I host – in terms of what time I tell to different guests!

The Law Of Comparative Advantage and Priorities

Over a decade ago I had written about two kinds of employees – those who offer “competitive advantage” and those who offer “comparative advantage”.

Quoting myself:

So in a “comparative advantage” job, you keep the job only because you make it easier for one or more colleagues to do more. You are clearly inferior to these colleagues in all the “components” of your job, but you don’t get fired only because you increase their productivity. You become the Friday to their Crusoe.

On the other hand, you can keep a job for “competitive advantage“. You are paid because there are one or more skills that the job demands in which you are better than your colleagues

Now, one issue with “comparative advantage” jobs is that sometimes it can lead to people being played out of position. And that can reduce the overall productivity of the team, especially when priorities change.

Let’s say you have 2 employees A and B, and 2 high-priority tasks X and Y. A dominates B – she is better and faster than B in both X and Y. In fact, B cannot do X at all, and is inferior to A when it comes to Y. Given these tasks and employees, the theory of comparative advantage says that A should do X and B should do Y. And that’s how you split it.

In this real world problem though, there can be a few issues – A might be better at X than B, but she just doesn’t want to do X. Secondly, by putting the slower B on Y, there is a floor on how soon Y can be delivered.

And if for some reason Y becomes high priority for the team, with the current work allocation there is no option than to just wait for B to finish Y, or get A to work on Y as well (thus leaving X in the lurch, and the otherwise good A unhappy). A sort of no win situation.

The whole team ends up depending on the otherwise weak B, a sort of version of this:

A corollary is that if you have been given what seems like a major responsibility it need not be because you are good at the task you’ve been given responsibility for. It could also be because you are “less worse” than your colleagues at this particular thing than you are at other things.

 

 

Lifting and arithmetic

At a party we hosted recently, we ended up talking a lot about lifting heavy weights in the gym. In the middle of the conversation, my wife wondered loudly as to why “so many intelligent people are into weightlifting nowadays”. A few theories got postulated in the following few minutes but I’m not going to talk about that here.

Anecdotally, this is true. The two people I hold responsible for getting me lift heavy weights are both people I consider rather intelligent. I discuss weights and lifting with quite a few other friends as well. Nassim Taleb, for a long time, kept tweeting about deadlifts, though now he has dialled back on strength training.

In 2012 or 2013 I had written about how hard it was to maintain a good diet and exercise regime. While I had stopped being really fat in 2009, my weight had started creeping up again and my triglyceride numbers hadn’t been good. I had found it hard to stick to a diet, and found the gym rather boring.

In response, one old friend (one of the intelligent people I mentioned above) sent me Mark Rippetoe’s Starting Strength (and a few other articles on cutting carbs, and high-fat diets). Starting Strength, in a way, brought back geekery into the gym, which had until then been taken over by “gym bros” doing bicep curls and staring into mirrors.

It’s been a long time since I read it, but it’s fascinating – I remember reading it and thinking it reminded me of IIT-JEE physics. He draws free body diagrams to explain why you should maintain a straight bar path. He talks about “moment arms” to explain why the bar should be over your mid-foot while deadlifting (ok this book we did discuss at the party in response to my wife’s question).

However, two incidents that happened last week gave me an idea on why “intelligent people” are drawn to lifting heavy barbells. It’s about challenging yourself to the right extent.

The gym that I go to (a rather kickass gym) has regular classes that most members attend. These classes focus on functional fitness (among other things, everyone is made to squat and press and deadlift), but I’ve for long found that these classes bore me so I just do my own thing (squats, press / bench and deadlift, on most days). Occasionally, though, like last Friday, I decide to “do the class”. And on these occasions, I remember why I don’t like the class.

The problem with the gym class is that I get bored. Most of the time, the exercises you are doing are of the sort where you lift well below capacity on each lift, but you do a lot of lifts. They train you not just for strength but also for endurance and metabolic conditioning. The problem with that for me is that because every single repetition is not challenging, I get bored. “Why do i need to do so much”, I think. Last Friday I exited the class midway, bored.

My daughter is having school holidays, and one of the things we have figured is that while she has grasped all her maths concepts rather soundly (the montessori system does a good job of that), she has completely failed to mug her tables. If I ask her what is “7 times 4” (for example), she takes half a minute, adds  7 four times and tells me.

Last Monday, I printed out (using Excel) all combinations of single digit multiplications and told her she “better mug it by Friday”. She completely refused to do it. There was no headway in her “learning”. I resorted to occasionally asking her simple arithmetic questions and making her answer immediately. While waiting to cross the road while on a walk, “what is six times eight?”. While waiting for the baker to give us bread “you gave him ?100 and the bread costs ?40. How much change should he give you?”. And so on.

She would occasionally answer but again her boredom was inherent. The concept learning had been challenging for her and she had learnt it. But this “repetitive practice” was boring and she would refuse to do it.

Then, last Friday, I decided to take it up a notch. I suddenly asked “what is four and a half times eight?” (she’s done fractions in school). This was a gamechanger.

Suddenly, by dialling up the challenge, she got interested, and with some prodding gave me the correct answer. An hour earlier, she had struggled for a minute to tell me what 8 times 7 is. However, when I asked her “what is eight times seven and a half?” she responded in a few seconds, “eight times seven is fifty six..” (and then proceeded to complete the solution).

Having exited my gym class midway just that morning, I was now able to make sense of everything. Practicing simple arithmetic for her is like light weight lifting for me. “Each rep” is not challenging in either case, and so we get bored and don’t want to do it. Dial up the challenge a little bit, such as bringing in fractions or making the weights very heavy, and now every rep is a challenge. The whole thing becomes more fun.

And if you are of the type that easily gets bored and wants to do things where each unit is challenging, barbell training is an obvious way to exercise. and “intelligent” people are more likely to get bored of routine stuff. And so they are taking to lifting heavy weights.

I guess my wife has her answer now.

 

Muggoos and overfitting

Back when I was a student, there was this (rather large) species of students who we used to call “muggoos”. They were called that because they would have a habit of “mugging up the answers” – basically they would learn verbatim stuff in the textbooks and other reading material, and then just spit it out during the exams.

They were incredibly hardworking, of course – since the volume of stuff to mug was immense – and they would make up for their general lack of understanding of the concepts with their massive memories and rote learning.

On average, they did rather well – with all that mugging, the downside was floored. However, they would stumble badly in case of any “open book exams” (where we would be allowed to carry textbooks into the exams) – since the value of mugging there was severely limited. I remember having an argument once with some topper-type muggoos (with generally much better grades than me ) on whether to keep exams in a particular course open book or closed book. They all wanted closed book of course.

This morning, I happened to remember this species while chatting with a friend. He was sending me some screenshots from ChatGPT and was marvelling at something which it supposedly made up (I remembered it as a popular meme from 4-5 years back). I immediately responded that ChatGPT was simply “overfitting” in this case.

Since this was a rather popular online meme, and a lot of tweets would have been part of ChatGPT’s training data, coming up with this “meme-y joke” was basically the algorithm remembering this exact pattern that occurred multiple times in the training set. There was no need to intuit or interpolate or hallucinate – the number of occurrences in the training set meant this was an “obvious joke”.

In that sense, muggoos are like badly trained pieces of artificial intelligence (well, I might argue that their intelligence IS artificial) – they haven’t learnt the concepts, so they are unable to be creative or hallucinate. However, they have been “trained” very very well on the stuff that is there in the textbooks (and other reading material) – and the moment they see part of that it’s easy for them to “complete the sentences”. So when questions in the exams come straight out of the reading materials (as they do in a LOT of indian universities and school boards) they find it easy to answer.

However, when tested on “concepts”, they now need to intuit – and infer based on their understanding. In that sense, they are like badly trained machine learning models.

One of the biggest pitfalls in machine learning is “overfitting” – where you build a model that is so optimised to the training data that it learns quirks of the data that you don’t want it to learn. It performs superbly on the training dataset. Now, when faced with an unknown (“out of syllabus”) test set, it underperforms like crazy. In machine learning, we use techniques such as cross validation to make sure algorithms don’t overfit.

That, however, is not how the conventional Indian education system trains you – throughout most of the education, you find that the “test set” is a subset of the “training set” (questions in examinations come straight out of the textbook). Consequently, people with the ability to mug find that it is a winning strategy to just “overfit” and learn the textbooks verbatim – the likelihood of being caught out by unseen test data is minimal.

And then IF they get out into the real world, they find that a lot of the “test data” is unknown, and having not learnt to truly learn from the data, they struggle.

PS: Overfitting is not the only way machine learning systems misbehave. Sometimes they end up learning the entirely wrong pattern!

Creative Grit!

This, by Annie Murphy Paul is a very interesting blogpost I came across today. This one is on “creative grit”.

There are two very interesting things that the blogpost talks about.

The first is that there are two ways to creativity – the “traditional way” (the way I’ve always seen it) is to think about the problem, internalise it and then somewhere “wait for inspiration to strike”.

The other method that the author talks about, referring to artist Franz Kline, is to “just keep trying”. Kline would make hundreds of paintings every night. And he would find that one (or few) of them would be good enough to work further on. So this form of creativity comes from repeated practice.

Then later in the blogpost, she also talks about some research on when creativity hits. Again, “traditionally” we are trained to think that if creativity has to “hit” us for a particular problem, that is much more likely to occur early on in our effort. Based on this, a lot of us creative people have come up with heuristics where if solution doesn’t occur within a few iterations of trying, we just give up and move on.

Annie Murphy Paul says that this is the incorrect approach. Quoting from her post:

Lucas and Nordgren call this the “creative cliff illusion”: we imagine that, after an initial upward leap, our creativity will then fall off a cliff—when in reality our creativity capacities are just getting ready to ascend.

We also misjudge the thoroughness of our search. In one study, people estimated that they had explored 75 percent of the solution space—when in fact they had covered only 20 to 30 percent of the relevant domain.

I sure should try the second method that she recommends – keep trying and occasionally you’ll be happy with the result. Or maybe I already do that with all my writing (this blog, my newsletters, etc.) – basically “spray and pray”. The reason I’ve managed to write so much is that I have a low bar for myself. So I write a lot of rubbish. And occasionally I end up writing something people like. On the other hand when I’m paid to write, I don’t “spray and pray”. And in trying to limit my downside I limit my upside as well.

And thinking about it, the reason this method works is that in creative pursuits only the wins matter. As long as you produce sufficient wins, no one cares about your duds!

While on the topic of creativity, here is an ancient lecture (maybe my first ever recorded “speech”) I gave on why “quality takes time”. This clearly shows that at least as of mid-2004 (coincidentally just before I started this blog) I used to strongly believe in the “wait for inspiration to strike” model of creativity.

Oh and btw, read the whole post. It’s worth it.

Paying doctors

Back in 2011-12, when I was about to go freelance, a friend told me about a simple formula on how I should price my services. “Take your expected annual income and divide it by 1000. That will be your hourly rate”, he said. I followed this policy fairly well, with reasonable success (though I think I shortchanged myself in some situations by massively underestimating how long a task would take – but that story is for another day).

The longer term effect of that has been that every time I see someone’s hourly rate, I multiply it by 1000 to guess that person’s approximate annual income (the basis being that as a full time worker, you “bill” for 2000 hours a year. As a freelancer you have “50% utilisation” and so you work 1000 hours).

And one set of people who have fairly transparent hourly rates are doctors – you know the number of appointments they give per hour, and what you paid for that, and you can back calculate their annual income based on that. The interesting thing is, for most doctors I’ve seen, based on this metric, what they earn for their level of eduction and years of experience seems rather low.

“So how do doctors earn?”, I wonder. Why is it still a prized profession while you might have a much better life being an engineer, for example?

Now you should remember that consultations are only one income stream for doctors. Those that practice surgery as well have a more lucrative stream – the hourly rates for surgeries far exceeds hourly rates of consultation. And so surgeons make far more than what I impute from what I’ve paid them for a consultation.

One possible reason for this arbitrage is the way insurance deals are structured – at least in India, out patient care is seldom paid for by insurance. As a consequence, hospitals and doctors cross-subsidise consultations with surgeries. They are able to get away with higher rates for surgeries because insurers are bearing the cost. Consultations, where patients generally pay out of their own pockets, are far more elastic.

This, however, leads to a problem for doctors who don’t do surgeries. Psychiatrists, for example. If they have to make money solely through consultations, their hourly rate must be far higher than that of doctors who also do surgeries. But then, is the market willing to bear this cost?

Now, I’m getting into conspiracy theory mode. If the amount non-surgeon doctors make is limited (thanks to market dynamics), the only way they can make sure they earn a decent living is by limiting supply. Could this be one reason India is under-supplied in a lot of non-surgical doctors? Again this is pure pure speculation, and not based in any fact.

Continuing with conspiracy theories, even for doctors who are surgeons, the only way to make a certain income is to have a threshold on the ratio of surgeries to consultations. And if this ratio (surgeries / consultations) goes too low, the doctors’ income suffers. Again, hippocratic oath aside, do hospitals try to game this metric, based on the current incentives?

On a more serious note, this distortion in the hourly earnings for surgeries versus consultations is one reason that India is also undersupplied with good general practitioners (GPs). Because GPs don’t do surgeries (though the Indian system means they are all licensed to perform surgeries, to the best of my knowledge), their earning potential is naturally capped. So the better doctors don’t want to be GPs.

How can we fix this distortion? How can we make sure we have better GPs? Insurance cover for outpatient care is one thing, but I’m not sure it is the silver bullet I’ve been making it out to be (and it will come with its own set of market distortions).

This entire post is me shooting from my hip. So please feel free to correct me iff I’m wrong.

Intelligent and Diligent

For whatever reason, when I was a schoolboy and first learnt of the word “diligent”, I assumed that it should be the opposite on intelligent. “Only people who are not intelligent need to be diligent”, the young I had reasoned.

And nearly thirty years later, I came across this stellar 2×2 on intelligence and diligence. I’ve read it in many places now, but will link to the version on farnam street blog. I’m copying this quote from the blog, which is apparently credited to two different military officers.

I divide my officers into four groups. There are clever, diligent, stupid, and lazy officers. Usually two characteristics are combined. Some are clever and diligent — their place is the General Staff. The next lot are stupid and lazy — they make up 90 percent of every army and are suited to routine duties. Anyone who is both clever and lazy is qualified for the highest leadership duties, because he possesses the intellectual clarity and the composure necessary for difficult decisions. One must beware of anyone who is stupid and diligent — he must not be entrusted with any responsibility because he will always cause only mischief.

Maybe I was up to something interesting back in the 1990s, even if it was rather self-serving. And maybe it is this concept I reprised in the late 2000s when I came up with “studs and fighters“. It was possibly my irritation with the “stupid and diligent” variety.

Now I’m thinking of this “stupid and diligent” 2×2 in terms of our schooling and education. Maybe there is this general feeling among parents, teachers and suchlike that intelligence is something you are “born with”, and you cannot become intelligent.

So the moment they spot a kid who is stupid and lazy, they decide that the best way to “improve” this kid is to make him/her more diligent, rather than more intelligent. In the short run this might work, since the kid is now able to do better in the school exams (which is what most teachers are optimising for). The long run effect, though, is that the kid, instead of ending up in the numerous but harmless “general staff” (stupid and lazy), ends up in the seemingly more competent but actually “dangerous, and only causing mischief” stupid and diligent quadrant.

In other words, our general schooling makes our adult population much more dangerous!

When Institutions Decay

A few weeks back, I’d written about “average and peak skills“. The basic idea in this blogpost is that in most jobs, the level of skills you need on most days (or the “average skill” you need) is far far inferior to the “peak skill” level required occasionally.

I didn’t think about this when I wrote that blogpost, but now I realise that a lot of institutional decay can be simply explained by ignoring this gap between average and peak skills required.

I was at my niece’s wedding this morning, and was talking to my wife about the nadaswara players (and more specifically about this tweet):

“Why do you even need a jalra”, she asked. And then I pointed out that the jalra guy had now started playing the nadaswara (volga). “Why do we need this entire band”, she went on, suggesting that we could potentially use a tape instead.

This is a classic case of peak and average skill. The average skill required by the nadaswara player (whether someone sitting there or just operating a tape) is to just play, play it well and play in sync with the dhol guy. And if you want to maximise for the sheer quality of the music played, then you might as well just buy a tape and play it at the venue.

However, the “peak skill” of the nadaswara player goes beyond that. He is supposed to function without instruction. He is supposed to keep an eye on what is happening at the wedding, have an idea of the rituals (given how much the rituals vary by community, this is nontrivial) and know what kind of music to play when (or not play at all). He is supposed to gauge the sense of the audience and adjust the sort of music he is playing accordingly.

And if you consider all these peak skills required, you realise that you need a live player rather than a tape. And you realise that you need someone who is fairly experienced since this kind of judgment is likely to come more easily to the player.

The problem with professions with big gaps between average and peak skills, and where peak skills are seldom called upon, is that penny-pinching managers can ignore the peak and just hire for average skill (I had mentioned this in my previous post on the topic as well).

In the short run, there is an advantage in that people with average skills for the job are far cheaper than those with peak skills for the job (and the former are unlikely to suffer motivational issues as well). Now, over a period of time you find that these average skilled people are able to do rather well (and are much cheaper and much lower maintenance than peak skilled people).

Soon you start questioning why you need the peak skill people after all. And start replacing them with average people. The more rare the requirement of peak skill is in the job, the longer you’ll be able to go on like this. And then one day you’ll find that the job on that day required a little more nuance and skill, and your current team is wholly incapable of handling it.

You replace your live music by tapes, and find that your music has got static and boring. You replace your bank tellers with a combination of ATMs and call centres, and find it impossible to serve that one customer with an idiosyncratic request. You replace your software engineers with people who don’t have that good an idea of algorithmic theory, and one day are saddled with inefficient code.

Ignoring peak skill required while hiring is like ignoring tail risk. Because it is so improbable, you think it’s okay to ignore it. And then when it hits you it hits you hard.

Maybe that’s why risk management is usually bundled into a finance person’s job – if the same person or department in charge of cutting costs is also responsible for managing risk, they should be able to make better tradeoffs.

Biennale!

I’m starting to write this at the beginning of today, as we go through the biennale.

We started with one place near the main venue where some volunteers has made some art. Some fairly trippy stuff, and some risqué stuff

Now on to today’s pertinent observations

  • This is only the fifth edition of the biennale
  • I quite love the artwork I’ve seen so far. And if not for the school I don’t think I would’ve seen all this art at all
  • Some of the art is “vaguely familiar”. And that intrigues me. The familiarity draws me in. The vagueness makes me want to keep seeing more of it. Sidhus quote about the bikini comes to mind
  • The biennale has this concept of “art mediators”. Effectively tour guides who explain the art and concepts around it. Our guide today was Safa. The comment I made about tour guides yesterday doesn’t apply to her. I’m enjoying her commentary.
  • Just now I heard another art mediator explain a piece of art that Safa had just explained. And the two are nearly orthogonal! I guess the thing with art is that it’s in the eyes of the beholder, or maybe the mediator
  • From the biennale venue (aspinwal house) you can see the kochi container port. To me, watching container ships getting loaded and unloaded is also art
  • Ok I finally have a hypothesis on what I consider as good art – something that compels me to keep looking at it. It could be stuff that is hard to interpret. It could be stuff with several dimensions. It could be things that tell a new story every time you look at them. That is the kind of art I like.
    • And so I like Paul Fernandes – so much going on in his art. And why I cut out a page from the times of india on Republic Day and pinned it on my wall
    • And so I like Picasso – it is so hard to interpret what he has done and there are so many dimensions I want to keep looking at it
    • If it gets too abstract though there is no “handle” to latch on to. And so it becomes hard to interpret
    • So far I haven’t been impressed by stuff with too much “messaging” – the message by definition makes it one dimensional and not something I want to keep seeing!
  • Art is fundamentally a “low precision low recall” activity. You can never see anywhere close to all the good art in the world (hence low recall)! Low precision because to find any good art you need to also go through a lot of art you can’t appreciate.

Ok now it’s afternoon and we’re at lunch. The post is long enough as it is and we’re done with the first session of the biennale. And I’m rather proud of the last two things I’ve mentioned here. So I’ll stop here.