Don’t want to humble-brag (but already right if you think about it what am I doing?) but this was the first hit in my search for a bit more on the question I asked at the end of an interesting talk today at SCU by Vivek Krishnamurthy, and it was exactly my question. Glad to know I am not in the far-away rafters when it comes to these issues.
Giving algorithms a sense of uncertainty could make them more ethical Posted on February 5, 2019 by Michael Rowe The algorithm could handle this uncertainty by computing multiple solutions and then giving humans a menu of options with their associated trade-offs. Say the AI system was meant to help make medical decisions. Instead of recommending one treatment over another, it could present three possible options: one for maximizing patient life span, another for minimizing patient suffering, and a third for minimizing cost. “Have the system be explicitly unsure and hand the dilemma back to the humans.” Hao, K. (2019). Giving algorithms a sense of uncertainty could make them more ethical. MIT Technology Review.
Source: Giving algorithms a sense of uncertainty could make them more ethical – /usr/space
I think about clinical reasoning like this; it’s what we call the kind of probabilistic thinking where we take a bunch of – sometimes contradictory – data and try to make a decision that can have varying levels of confidence. For example, “If A, then probably D. But if A and B, then unlikely to be D. If C, then definitely not D”. Algorithms (and novice clinicians) are quite poor at this kind of reasoning, which is why they’ve traditionally not been used for clinical decision-making and ethical reasoning (and why novice clinicians tend not to handle clinical uncertainty very well). But if it turns out that machine learning algorithms are able to manage conditions of uncertainty and provide a range of options that humans can act on, given a wide variety of preferences and contexts, it may be that machines will be one step closer to doing our reasoning for us.
One of the most controversial of these measures is SB 50. Hailed by advocates as a solution to the Bay Area’s housing shortage, the bill would override cities’ density rules, height limits and parking requirements in areas near public transit hubs. For example, projects within a half-mile of major transit stops — including two Caltrain stations in Palo Alto and one on the border with Mountain View — could be up to 45 feet tall, or about four stories. About 7,000 parcels, or 40 percent of Palo Alto’s total parcels, would be subject to SB 50 rules — enough to transform Palo Alto from a community of predominantly single-family homes into a city dominated by townhouses and apartments, according to the Embarcadero Institute.The legislation could cause the city’s population to grow to 2.7 times its current size and bring up to 30,000 new students to Palo Alto, the report said, potentially stretching the capacity of the local schools in a city renowned for the quality of its public education system. SB 50 also could bring as many as 90,000 additional vehicles to town, according to the report.
Source: Report: Housing bill law could almost triple size of Palo Alto
I have been slacking. For Christmas I got several novels.
Children of Time by Adrian Tchaikovsky. Excellent big sci-fi opera, with genetic engineering and big questions.
The Peace War by Vernor Vinge. I started skeptical but got drawn in. By the end I really enjoyed it.
Marooned in Realtime by Vernor Vinge. I had read it years ago, and since it is a sequel of shorts to The Peace War I had to skim-read it again. The bobbles were one of the great “discoveries” in sci-fi.
Pnin by Vladimir Nabokov. Reading for our neighborhood book club. So inventive, so much fun.
Radical Markets by Eric Posner and E. Glen Weyl. Read it with Economics students in a discussion group. Lots of food for thought, but a bit maddening as they are sloppy in not thinking through many obvious questions. But always worth a reminder that institutions that seem solid and right are contingent and changeable.
The Story of a New Name by Elena Ferrante. it was gripping the whole way through. but I will admit I got stuck halfway through the third volume. let’s say I am saving it for a rainy day!
The Silence of Our Friends by Mark Long, Jim Demonakos, and Nate Powell. A graphic novel. Compelling, if in the end a bit thin.
Cloudsplitter by Russell Banks. Worth the slog. If only to march through the amazing history of John Brown via wikipedia.
I saw the book in the library. I hesitated. I seemed to recall it had been received with coolness by reviewers? But I thought to myself, he is such a good writer, can it really be that bad? And so I started reading last week, and pretty much looked forward to the end of the day, each day, so I could read another 50 pages. Wonderful prose. A brilliant “British” novel. Naturally, you have to be steeped in the British writing tradition to really enjoy, but if you are, you will. More importantly, Ishiguro introduces a writing technique that was largely new to me: repetition, haziness, circling back, mixing dialogue with interior monologue, unfinished tales. On this last, I have read many “original” folktales, the kind collected by the early ‘anthropologists’ who sat with loquacious storytellers in villages all across Africa. Those unvarnished folktales have a lot in common with the novel. They are often seem unfinished, to the modern reader, and they often seem to drift from anecdote to anecdote without a common thread. Part of that, presumably, is that they were oral tales quasi-made up in the course of interrupted conversation, with some audience participation. This novel evokes that feeling: the hesitations almost beg for one of the characters to start filling in a few details, or to change the direction of the story. And that sometimes happens. The characters correct each other: “No, that is not what happened, what happened was…” Characters that are introduced, and seem important, disappear into the mist.
Aside from the writing and technique, the novel provokes a profound meditation: what is this narrative of our lives, individual and collective, when so much is forgotten and invented? Ishiguro does not neatly answer the question, because the novel is about the asking. So you have to be comfortable with ambiguity.
Neil Gaiman wrote a much better review than I ever could for The New York Times.
I have not followed Sudan politics in over a decade, but I do follow Burkina Faso pretty closely, and in October 2014 long-time president Blaise Compaoré was ousted by street protests. Basically, regime insiders had to choose when to run, and as more ran, Compaoré himself decided to run, and insiders who remained took the reins of power in collaboration with civilian leaders in a long one year transition marked by episodic violence. Here are a number of things to consider for a transition in Sudan, which shares many similarities (and one notable difference: much more extensive militarization of society resulting from decades of protracted ‘frontier’ wars, so some kind of disarmament might be considered, though that has been a thorny issue for South Sudan). Here’s my two cents of perspective for civilian-rule leaders, gleaned from my understanding of Burkina Faso’s experience.
- Get dangerous insiders out of proximity to power right away. The biggest threat to the transition in Burkina Faso came in September 2015 when Gilbert Diendere, a top man in the formerly ultra-powerful presidential guard, staged a coup. He almost got away with it, but most of the regular army sided with the civilians. He and his coup comrades are now on trial. The disbanding of the presidential guard was one of the most consequential decisions the transition faced. It will always be risky. I would deal with it early and send the leaders to The Hague where they can use their wealth to hire elite lawyers.
- Don’t worry about legality and constitutionalism. The transition leaders in Burkina (whether deliberate or not) let the issue of “what exactly is the legal status of our state) not bother them too much. The key I think is having a supreme constitutional council that will be ultimate arbiter of legality that is stacked with civilian rule promoters who will interpret the contradictory thicket of non-legality in ways that will promote consolidation of civilian rule and rule against previous regime insiders clinging to power by appealing to previous regime technicalities. The biggest issue for this court in Burkina Faso was to determine whether old regime members could participate in the elections. The court ruled in the negative.
- Make a list of a projects and get them done expeditiously. Government is the largest employer and its multiplier is huge. A transition needs to be immediately giving out public works contracts and reducing the disruptive impact of the transition.
- Let the media flourish. For two reasons. One is that a competitive media probably fractionalizes potential opposition to peaceful civilian rule. Spoilers need to form a coalition to regroup, but if the former regime coalition is permitted to communicate, chances are they will be less likely to plot in secret. Another is that the free media is actually a significant employer of well-educated young people, and so gives them a bigger stake in promoting a free society. An excellent review by Nael Jebril, Václav Stetka, Matthew Loveless, however, suggests there is no robust academic basis for my suggestions 😉
I am not yet even close to Jennifer Weiner’s level after two years of one hour a week lessons, but I know the feeling (and the 15 year old).
I open my book to Chopin’s waltz in D flat major, the “Minute Waltz,” so called because you’re meant to play it in under a minute. Right now, I’m averaging around five. As I start the first trill my 15-year-old daughter, who these days speaks mostly in sarcasm, strolls by. Sometimes she’ll do a mocking balletic leap as I play, or just emphatically shut her bedroom door. Tonight, she does neither. “Hey, Mom,” she says, “that was really good!”It wasn’t. It was just O.K. But that is good enough.
The opinion piece is here.
The Power, by Naomi Alderman. Thrilling in its conceit and execution. Stumbles towards the end (and what ambitious book doesn’t). Nice to see an interesting short frame device (the letters) in popular fiction.
News of the World, by Paulette Jiles. It is a simple and almost maudlin story, shades of and homage to True Grit, but the violent reality of 1870s Texas leaves you feeling like each turn of the page has plucked a hair off your head and turned two more gray.
Eileen, by Ottessa Moshfegh. So cranky, so different. The prose makes you keep reading even as the throbbing gristle of a book makes you want to put it down.
Artemis, by Andy Weir. Almost unreadable. After his success with The Martian, you can just see his agent telling him to write a star vehicle for Jennifer Lawrence or Emily Blunt. Basically a screenplay for a bad Hollywood movie. With lots of explosions and cliffhangers. Oh look, I was right, the movie deal was in the works even before the book came out.