Mind and Iron: Why We Should Embrace Robot Umpires
Also what's this whole quantum computing thing about?
Hi and welcome back to another saucy episode of Mind and Iron. I'm Steven Zeitchik, veteran of The Washington Post and Los Angeles Times and lead confectioneer at this journalistic candymaker.
Every Thursday we bring you the highlights of tech news as it's impacting our lives. No spin, no corporate-media agendas. No cult of moguldom either. Please consider supporting our independent mission.
We here at M&I headquarters had a death in the family this week. So we're going to go a little shorter but not leave you hanging. Also and relatedly, I wanted to dedicate this issue to Uncle N, a true gem of a human being and icon of teaching who most certainly did not know much about AI but weirdly had a digital organizer many, many years before anyone else did. (And then kept said organizer into the iPhone era.) This one's for you.
On tap this week: the matter of robo-umps, as it's been in the news a lot lately with spring training getting going. Technically even the term robo-ump is a misnomer, as R2-D2 is decidedly not standing up the first-base line. We'll clarify what's happening and offer a surprising take.
Also, you're probably hearing about quantum computing, particularly a discovery that Microsoft just announced. Right now quantum computing is a neutral term, perhaps an opaque one. It may not be for long, and may supercharge us into the next era, even leaving the AI one in the dust. A quick primer on this whole qubit thing.
First, the future-world quote of the week.
“Can we just be judged by humans?"
— Toronto Blue Jays pitcher and future Hall of Famer Max Scherzer, in perhaps the first locker-room soundbite that doubles as a techno-philosophy manifesto
Let's get to the messy business of building the future.
IronSupplement
Everything you do — and don’t — need to know in future-world this week
Machines judging baseball; machines changing physics
1. THE IDEA HAS BEEN CREEPING TOWARD OUR SPORTS WORLD for a while now: robo-umps. More specifically, ABS, the "Automatic Balls-Strikes" system that began when a Danish company started tracking ball movements for golf swings then realized there was money lurking under them home plates.
Making a deal with the independent Atlantic League back in 2019, the company and its product started inching closer to Major League Baseball ever since, beginning with the low minor leagues in the spring of 2021 and steadily working its way up the chain in the seasons that followed.
The idea is simple. The tech measures a standard-issue strike zone based on the shape of home plate and customizes it to each player based on their body types. Then it measures where a pitch went relative to a given player and makes a call with an accompanying graphic.
This is not a hard feat given modern sensors and camera tech, as anyone who has watched the Hawk-Eye system in tennis knows. The out calls at the Australian and U.S. Opens (and this coming summer for the first time Wimbledon) are now fully automated. And with that change the wrong calls that in the past have decided matches (and the McEnroe-esque yelling that has interrupted them) are gone.
ABS is basically an attempt to standardize what has historically been very non-standard, with different umps calling different strike zones (and sometimes the same ump calling different strike zones within the game). The ump still stands there — they're needed for plays at the plate — but they’re not in control of balls and strikes, instead just reciting what the system tells them in their ear.
This year ABS is getting closer to Major League teams than it ever has before — 19 clubs, all playing in Florida, are using it for their spring training games. And one small problem in our big complicated world would seem solved, at least for a few weeks in the tropical heat.
As this is Major League Baseball, though, caution is the better part of logic, and so they're not really trying out the system. They're keeping the human umps calling balls and strikes right where they are and giving teams up to two incorrect ABS challenges per game. When a pitcher, catcher or batter doesn't like a call, he can challenge the ump. Then the graphic of the pitch flashes on the scoreboard and the call is either overturned or upheld. Here’s an example of it being used in an exhibition game a few years back.
So far it seems to be working just fine — accuracy is high, players are figuring out when to use the challenges (with one exception; more on that in a second), and thus the latest gateway to a potentially full-time ABS system in the regular-season 162 has been crossed. Should it get there? That's stickier.
On its face this situation would seem very much akin to what happened with those tennis Grand Slams, which also used the challenge system for a while before switching over to full automation. The idea of using a review system for cases of millimetric accuracy is an interesting one philosophically and doesn't entirely cohere — it basically says that we trust the tech, it's totally foolproof, but for some reason we're not ready to give up on those fallible humans unless someone on a given play is really mad about it.
In tennis that really made no sense because the line doesn't change based on the match or the player. It's entirely set. There is no art, as it were, to making a line call; you're either getting it wrong or right. As Hawk-Eye has shown to always get it right, go with the machines full-time.
Baseball, in its quaint anachronism, offers a different proposition. There is of course no such thing as a fully accurate strike call — each umpire from the beginning of time has seen the strike zone differently. And even when they haven't, catchers are known — indeed, celebrated — for "framing" a pitch. Which is a fancy way of saying fool the umpire into thinking something's a strike.
Baseball purists argue this is a beautiful part of the game that shouldn't be eliminated, this whole human persuasion aspect, and if we have to sacrifice a little bit of accuracy or fairness to achieve that, so be it. I think this is what Scherzer was positing when he said "can't we just be judged by humans," though he may have just been annoyed ABS called a ball.
(Actually he had a much longer spiel in which he bizarrely seemed to miss the basic math of this. Talking about how some pitches called balls will become strikes upon challenge and some strikes balls, he said, "my gut is that it’s going to be 50-50 and you don’t really have a net gain." Or, like, it'll be a 100 percent gain since the goal is more correct calls, not all strikes or all balls? He also said, "Do we really need to disrupt the game? I think humans are defined by humans.” People are people so why should it be, your AI should make me feel so awfully.)
Anyway, by now if you're a regular reader you know we tend to side with the humans on these questions. Soul is kinda important in this increasingly techno world, and if we have to lose a little accuracy to acquire that soul, well, that's a currency worth expending.
So it may surprise you to hear that I’m siding with the machines on this one. Bring on ABS. The sooner the better.
The reason is not because the human strike zone is such a travesty. The truth is Scherzer is right on this point. Pro baseball has been around for 150 years without the technology, and it did just fine. Yes, it led to some yelling and disagreements, and even some teams and players getting jobbed. It also led to a cool art of persuasion from pitcher and catcher (and batter), and gave humans an important job they hopefully didn't abuse. Also, all those cool nose-to-nose umpire-manager fights where what they were really talking about was where they were going to dinner.
On balance, that's a wash. We're not losing THAT much rejecting ABS.
But that's within the game of baseball. The problem with such a fierce opposition to robo-umps goes beyond that, to a larger social question.
One of the biggest challenges as many of us seek to moderate a Big Tech that wants to develop every innovation at breakneck speed and then impose it on us is making a coherent counter-argument. Accelerationism, after all, has a powerful rhetorical weapon on its side: human existence has been marked by technology at every turn since the dawn of fire, and look at how that's all turned out. If you could snap your fingers and make the cell phone in your pocket disappear, would you? So trust us and let us do our thing, they say.
It's a powerful argument. And one that can be countered by making surgical cases against particular technologies where the bad documentably outweighs the good. Very specific, well-evidenced cases. And even then. We’re going to have a hard enough time making a more human world as it is given the tide of history/these companies; we don't need to invent new enemies. Resisting all technological change just for the sake of it will weaken the argument in the times we really need it.
Like, adopting AI to serve as our kid's companion or as the repository of or memories; that’s a more genuine hazard. That's the kind of stuff we should be fighting. Not a set of cameras that can clearly see a 100 mph fastball better than any human.
In short, let's choose our battles. Saying every technological enhancement is bad — especially when it's clear as day making a game fairer — means the plea falls on deaf ears when it really is bad. Let's not cry wolf when it might just be a hologram.
Ditto on the job displacement front. I'm all for arguing that many professions need humans and the rush to replace them with machines is problematic and misguided. An AI shouldn't replace an actor because an actor can bring a humanity that AI never could. But if we start saying that an AI shouldn't replace a plate-measurer, then we're cheapening the whole displacement argument. Then we're not saying humans bring something valuable. We're saying protect their work even when they don't. And once that happens, not only will it not fly (companies are never going to spend more money for something less valuable) but it’ll make people inured to listening when the humans really do add something valuable.
So bring on full ABS. Balls and strikes will now be clear-cut, no more fighting. And if you really miss some baseball-diamond violence, give Max Scherzer a call. I'm sure he'll be happy to meet you in the locker room and trash a few laptops with you.
2. WHEN ONE THINKS OF AMAZON, ONE THINKS OF REALLY GOOD DEALS ON PAPER TOWELS, NOT SCIENTIFIC JOURNAL ARTICLES WITH TITLES LIKE “HARDWARE-EFFICIENT QUANTUM ERROR CORRECTION VIA CONCATENATED BOSONIC QUBITS.”
Yet welcome to the upside-down, as this week we got just that, Amazon publishing such a paper in Nature ahead of debuting a new chip called Ocelot. The company is firing back after Microsoft announced its own quantum chip last week, going with its own article in Nature too. Their title? “Interferometric single-shot parity measurement in InAs–Al hybrid devices.” Sometimes you just gotta go for the clickbait.
Your first thought might be, man, these tech companies will fight over anything, won’t they? But your second thought might be — so this seems like something really worth fighting about, eh?
Quantum computing can fill a book or three, not a Substack, and certainly a lot of copies of Nature. So here’s the extremely, extremely quick version.
Right now traditional computing uses data packets, which expresses language via bits that are either on or off, ones and zeroes — it’s literally a binary system. But with the help of quantum mechanics and subatomic particles, quantum computing’s “qubits” can be in both states at the same time. Which makes things go faster…a lot faster.
So strange is the power of qubits is that it cannot be apprehended — literally. The physics-based simultaneity of it means it can’t really be seen the way another object would. As a NY Times piece once famously noted, “If any of us actually looked at the fundamental building block of a quantum computer, called a qubit, it would cease to be a qubit.” (I know.)
The best (understandable) analogy I’ve heard for quantum computing vs the traditional kind is the requirement to move linearly across city blocks as opposed to passing through buildings with quasi-superhuman powers. Not only would you get where you’re going faster, but your whole way of moving would change.
Or, in this case, your whole method of computing. If quantum computing were to be achieved, you could (theoretically) solve problems no classic computer could. You could (theoretically) crack cryptographic codes and (theoretically) achieve new levels of drug-discovery breakthroughs, because you (theoretically) wouldn’t be bound by the same rules of computing. I say theoretically, because nobody knows what quantum computing can do, as it’s yet to be fully invented let alone tested. This is very experimental stuff.
What we do have are some promising intermediate steps to that steroidal end goal. That’s what Microsoft announced last week. What they did was claim something called “topological qubits” that create those qubits by keeping materials in a state of matter that is neither solid, liquid nor gas, thus enabling that simultaneity. Again, very theoretical, as even some of the scientists who’ve been briefed on the research are only cautiously optimistic. But still, the kind of vista that even several years ago was hard to imagine.
Nor is Microsoft the only company in the mix. In addition to Amazon’s first-ever quantum-chip prototype, IBM and Google are both are the fore too. In December Google announced its own chip, Willow, that could supposedly undertake a computation in five minutes what would take a traditional computer — wait for it — a septillion years to conquer. Which, for comparison purposes, is nearly as long as a Hollywood awards show.
So clearly there’s an arms race. What happens when we get to the end of it? What will all this promise of speed and power mean for how we live? That’s a thornier question, and one on which we need to wait. (Possibly a while — while the developments are coming much faster in recent years, mainstream usage could still be a decade or more away.) But the idea, at least, is that quantum computing can power everything from batteries to AI, enabling a supersizing of the world we’re about to head into in such a way it changes not just quantitatively but qualitatively as well.
The nature of that quality is hard to predict. But clearly we’re heading there. Just five years ago Google made news by achieving so-called quantum supremacy after scientists there built a computer that took only about three minutes to solve a problem that would have taken a classic machine 10,000 years. Compared to the septillions of today, it was practically a Commodore 64.
The Mind and Iron Totally Scientific Apocalypse Score
Every week we bring you the TSAS — the TOTALLY SCIENTIFIC APOCALYPSE SCORE (tm). It’s a barometer of the biggest future-world news of the week, from a sink-to-our-doom -5 or -6 to a life-is-great +5 or +6 the other way. This year started on kind of a good note. But it’s been a little rough since. This week? We’re on OK ground.
ABS OFFERS SOME TANGIBLE BENEFITS TO THE MACHINES: Why fight it?: +2.0
QUANTUM COMPUTING (THEORETICALLY) OFFERS SOME BENEFITS : +2.5