Franco Calabrese

[Franco Calabrese @Franco_Calabres]

@deanwball Oh, man, you've done it now. When you misspell a word in a prayer ("Lord, grant me the cajones...") the manifestation of divine outrage is sure to be devastating but may be delayed for years just to keep you in suspense.

[Franco Calabrese @Franco_Calabres]

@jxnlco I had the same reaction when I was sporadically drafted out of engineering to help with sales and marketing. The downside, IMO, is that there are incentives to shade the truth by selective emphasis if not outright deception even in the most principled companies.

[Franco Calabrese @Franco_Calabres]

@tylerblack32 Perfectly reasonable if his office sessions involve a lot of open wounds and splashing bodily fluids ;-)

[Franco Calabrese @Franco_Calabres]

@binarybits Excellent observation. For complicated tasks where the contours of details, sequence, and evaluation aren't obvious, directing others (humans or machines) successfully has always required skill and insight.

[Franco Calabrese @Franco_Calabres]

@yacineMTB I have to suspect that you've never had much personal experience with the consequences of poverty in large families.

[Franco Calabrese @Franco_Calabres]

@TylerSbrt @midjourney What a radical idea!

[Franco Calabrese @Franco_Calabres]

@steffingomes @_TomHoward I admire your restraint.

[Franco Calabrese @Franco_Calabres]

@Ducnghia16 Glen was good but so are many dozens of other good guitar players. Not to mention that nearly everybody in the orchestra behind is at least as technically adroit.

[Franco Calabrese @Franco_Calabres]

@paradite_ An idea doesn't have truly novel to have tremendous value. It simply has to be new to the observer and useful in context. That's a much lower standard and the same one we fervently hope we and our human assistants will meet every so often.

[Franco Calabrese @Franco_Calabres]

@nathanbenaich There's some point at which the availability of new technology justifies letting old skills wither. Don't know that we're there yet with (say) adenoma detection, but eventually it becomes a poor use of physician time to update skills that have declining value to patients.

[Franco Calabrese @Franco_Calabres]

@BuiltByJuice Thanks for the guesswork but I prefer actual evidence.

[Franco Calabrese @Franco_Calabres]

@rachel_l_woods Can't answer that, but I'll note that some months ago I started a knowledge base project using Obsidian as the back end. Within weeks the only Obsidian remnant was the file layout; Codex-generated software was doing all the reading and writing to my specifications.

[Franco Calabrese @Franco_Calabres]

@tylerblack32 @ScienceUpFirst In the early 1970s I hung out with a friend sweating through his psychiatry residency under Paul McHugh. We often bemoaned that too much confidence pushes you toward dogmatic foolishness while too little can be outright paralytic. No clearly demarcated balance point.

[Franco Calabrese @Franco_Calabres]

@ziv_ravid This is near the top of my list of worries as well. It portends more unnecessary wealth inequality and will fuel populist conflict. As AI avoiders and opponents lose economic clout they're more apt to focus on a sense of grievance than to put AI to work for themselves.

[Franco Calabrese @Franco_Calabres]

@AndyMasley For a couple of months I've been hip deep in a project to glean insights about what nudges achievement test scores up and down and it's dauntingly hard. Whole careers are built on staring at the numbers, applying every conceivable statistical manipulation, and still wondering.

[Franco Calabrese @Franco_Calabres]

@TheAhmadOsman I'm always a little uncomfortable with unqualified "never give in" advice. Courage and perseverance are great, but only when paired with a willingness to detect and acknowledge personal bias and fallibility. Lacking that you've just got a recipe for irremediable assholery.

[Franco Calabrese @Franco_Calabres]

@per_simmons_ Thanks. Great stuff.

[Franco Calabrese @Franco_Calabres]

@uniquehumann @AnthonyNAguirre I agree in principle, but I also suspect that in practice the number of important and erroneous systemic biases we absorb from uncritical acceptance of human opinions may swamp that effect for a very long time. Not sure whether or when there will be net loss of diversity.

[Franco Calabrese @Franco_Calabres]

@michaelzlin Really clever work. Rooting for subsequent testing to go well, too. Even knowing that the concept of "deserve" is badly misplaced here, I can't shake the feeling that you guys "deserve" a win with this.