Loved this whimsy.

I had a lovely hour last night chasing the Orion Nebula with my 16yo at a local soccer field with our new scope. We enjoyed seeing more things every time we swapped lenses, and describing what we saw. Her eyes are better, so she figured out first that the one blurry star in the wide view was actually a tiny trapezoid of four stars very close together. At higher power they separated quite nicely, and we began to see the contours of the nebula. Not bad for the local soccer field, a crescent moon straight overhead, and some high clouds. Plus, memories.

The scope is a used Telescopics 8" reflector from the 1980s in a pretty wooden cradle on a Dobsonian mount. The mirror seems to be really good. (Photo by Bob Parks from the club sale.)

I see crypto as dystopian by design, but Daniel Meissler’s re-imagining at least lets me glimpse why it might have general appeal.

Take a multi-player world like MineCraft where people create amazing fantastic, complicated things. That seems to be working fine right now, but it seems a really creative builder or gamer needs YouTube or Patreon or a side job. Meissler:

I’m not sure this is better but it’s not obviously bad. It has the benefits and flaws of how we build valuable things & experiences in the real world.

And I’m not sure it needs blockchain – didn’t Steam already do this a decade ago without crypto?

But at least he sketches why there might a coordination problem to solve, rather than a cool hammer hunting nails.

Today it was this 2016 Jacobs piece I found from his tribute to Paul Farmer.

I have at times been in groups of people who know and respect the work of Care Net, but if in those contexts I mention my admiration for the work of Paul Farmer and Partners in Health, I am liable to get some suspicious looks. …

In other groups, people join enthusiastically in my praise for Paul Farmer — but become nervous when I mention my admiration for Care Net. …

And yet both Partners in Health and Care Net are pursuing the Biblical mandate to care for the weakest and most helpless among us. In so many ways they are doing the same work, and even are dedicated to the same goal — the preservation and healing of the lives of people made in the image of God. Why must we see them in opposition to each other?

I signed into µ.blog to bookmark some tech piece and found instead that Paul Farmer has died.

Not just “if you write, give a damn”, but about us. This is about community and relationships. She’s tempted to outrage, but she lives there, and “outrage is where relationships go to die.”

Outrage doesn’t get curious about why. It mutes the Principle of Charity. It often reacts to imagined motive more than act. Witness Colin Kaepernick taking a knee.

To be sure, there is plenty of anti-Semitism about, and Maus is charismatic megafauna, worthy of defense. But,

The meeting minutes suggest that their objections to Art Spiegelman’s _Maus_ series had less to do with the subject matter and more to do with a purity narrative _that never seems to die_, no matter the zip code.

Around here, antisemitism tends to take far more flagrant forms,

The minutes show a board mostly committed to teaching the Holocaust, but deeply uncomfortable making main unit text contain, as Renki summarizes, “profanity, sexuality, violence and… a suicide scene.” They wanted to just fully redact eight words and an image, but were advised that might exceed “fair use”.

That’s like restricting Alice in Wonderland for promoting poison, but as Renki notes, purity culture is universal: my demographic doesn’t mind literary profanity, but we have put two prize-winning classics on the Top 10 Banned Books list, for violating other norms.

Back to Kimball Coe:

This seems so very Catholic:

Also worth reading, Will Wheaton’s response to a question about Joss Whedon.

Python Anti-Pattern: Default Mutable Arguments

@SourceryAI found a bad habit I didn’t know I had: def f(names: list=[]):.   This would be nearly impossible to debug.

docs.sourcery.ai/suggestio…

Software pushes us to get better as people, on her blog Jessitron.

From a recent AAAS member spotlight on Alan Alda. It continues:

Today, the Alda Center has thus far trained some 20,000 scientists and medical professionals around the world to share science in a clear, engaging and accessible manner.

Periodically a song from “The Greatest Showman” is on my mental background music.

I just noticed my brain had substituted for “anthem”:

 Let this promise in me start
 Like an anvil in my heart
 ...

🤔

Bookmark: Alan Jacobs excerpting a recent Rowan Williams essay in The New Statesman. Education, Science, and Humanities.

Dear Self: Remember periodically to reread Dorothy Bishop’s (@DeevyBee) recent reflection on humility and choosing battles, “Why I am not engaging with the Reading Wars”. 

deevybee.blogspot.com/2022/01/w…

Huh! Quite OK Image compression provides a simple fast algorithm and format that seems to benchmark well.  Encodes the next pixel as either run-length, small diffs, or recently-seen.

Headlines about the death of theory are philosopher clickbait. Fortunately Laura Spinney’s article is more self-aware than the headline:

(Note: Laura Spinney also wrote Pale Rider, a history of the 1918 flu.)

Big Data?

Forget Facebook for a moment. Image classification is the undisputed success of black-box AI: we don’t know how to write a program to recognize cats, but we can train a neural net on lots of picture and “automate the ineffable”.

But we’ve had theory-less ways to recognize cat images for millions of years. Heck, we have recorded images of cats from thousands of years ago. Automating the ineffable, in Kozyrkov’s lovely phrase, is unspeakably cool, but it has no bearing on the death of theory. It just lets machines do theory-free what we’ve been doing theory-free already.

Mis-understanding

The problem with black boxes is supposedly that we don’t understand what they’re doing. Hence DARPA’s “Third Wave” of Explainable AI. Kozyrkov thinks testing is better than explaining - after all we trust humans and they can’t explain what they’re doing.

I’m more with DARPA than Kozyrkov here: explainable is important because it tells us how to anticipate failure. We trust inexplicable humans because we basically understand their failure modes. We’re limited, but not fragile.

But theory doesn’t mean understanding anyway. That cat got out of the bag with quantum mechanics. Ahem.

Apparently the whole of quantum theory follows from startlingly simple assumptions about information. That makes for a fascinating new Argument from Design, with the twist that the universe was designed for non-humans, because humans neither grasp the theory nor the world it describes. Most of us don’t understand quantum. Well maybe Feynman, though even he suggested he might not really understand.

Though Feynman and others seem happy to be instrumentalist about theory. Maybe derivability is enough. It is a kind of understanding, and we might grant that to quantum.

But then why not grant it to black-box AI? Just because the final thing is a pile of linear algebra rather than a few differential equations?

Theory-free?

I think it was Wheeler or Penrose – one of those types anyway – who imagined we met clearly advanced aliens who also seemed to have answered most of our open mathematical questions.

And then imagined our disappointment when we discovered that their highly practical proofs amounted to using fast computers to show they held for all numbers tried so far. However large that bound was, we should be rightly disappointed by their lack of ambition and rigor.

Theory-free is science-free. A colleague (Richard de Rozario) opined that “theory-free science” is a category error. It confuses science with prediction, when science is also the framework where we test predictions, and the error-correction system for generating theories.

Back to the article

Three examples from the article:

Machines can predict better than professionals.

Certainly. Since the 1970s when Meehl showed that simple linear regressions could outpredict psychiatrists, clinicians, and other professionals. In later work he showed they could do that even if the parameters were random.

So beating these humans isn’t prediction trumping theory. It’s just showing disciplines with really bad theory.

Prospecting Gaps

I admire Tom Griffiths, and any work he does. He’s one of the top cognitive scientists around, and using neural nets to probe the gaps in prospect theory is clever; whether it yield epicycles or breakthroughs it should advance the field.

He’s right that more data means you can support more epicycles. But basic insights Wallace’s MML remain: if the sum of your theory + data is not smaller than the data, you don’t have an explanation.

Regularization

AlphaFold’s jumping-off point was the ability of human gamers to out-fold traditional models. The gamers intuitively discovered patterns – though they couldn’t fully articulate them. So this was just another case of automating the ineffable.

But the deep nets that do this are still fragile – they fail in surprising ways that humans don’t, and they are subject to bizarre hacks, because their ineffable theory just isn’t strong enough. Not yet anyway.

So we see that while half of success of Deep Nets is Moore’s law and Thank God for Gamers, the other half is tricks to regularize the model.

That is, to reduce its flexibility.

I daresay, to push it towards theory.

Marvel’s The Eternals is not my favorite, but I’m liking it a bit more after Film Theory linked the plot to Thanos – and gave the first plausible Thanos redemption I’ve heard.