First off, though this posting just isn’t total political… I’ll provide a warning to you activists on the market.
Whereas I feel protest marches are among the many least efficient sorts of resistance – (particularly since MAGAs reside for one factor: to drink the tears of each smartypants skilled/scienctist/civil-servant and so forth.) — I nonetheless reward you lively of us who’re preventing nonetheless you possibly can for the Nice (now endangered) Exxperiment. Nonetheless could I level out how deeply silly the organizers of this 50 501 Motion are?
There’s at all times extra about AI – and therefore a set of hyperlinks to…
== The AI dilemmas and myriad-lemmas proceed ==
I’ve been amassing a lot materials on the subject… and extra retains pouring in. Although (alas!) so little of it enlightening about easy methods to flip this tech revolution right into a constructive sum end result for all.
Nonetheless, let’s begin with a potpourri…
1. A brand new research by Caltech and UC Riverside uncovers the hidden toll that AI exacts on human well being, from chip manufacturing to knowledge middle operation.
2. And likewise this re: my alma mater: Caltech researchers have developed mind–machine interfaces that may interpret knowledge from neural exercise even after the sign from an implant has grow to be much less clear.
3. Swinging from course of to implications… my buddy from the UK (by way of Panama) Calum Chace (creator of Surviving AI: The Promise & Peril of Synthetic Intelligence) despatched me this one from his Conscium Undertaking re: “Ideas for Accountable AI Consciousness Analysis”. Whereas detailed and illuminating, the work is imprecise about a very powerful issues, like easy methods to inform ‘consciousness’ from a system that feigns it… and whether or not that even issues.
Furthermore, not one of the authors cited right here refers to how the subject was pre-explored in science fiction. Take the dive into “what’s consciousness?” that you simply’ll discover in Peter Watts’s novel “Blindsight.”
…whereby Watts makes the case {that a} sense of self just isn’t even mandatory to ensure that a being to behave in methods which might be actively clever, communicative and even ferociously self-interested.
All you want is evolution. And an total system wherein evolution stays (as in nature) zero-sum. Which – I maintain making an attempt to inform of us – just isn’t essentially fore-ordained.
== And but extra AI associated Miscellany ==
4. Augmented actuality glasses with face-recognition and entry to world databases… now the place have we seen this earlier than? How about in nice element in Existence?
5. On the MINDPLEX Podcast with AI pioneers Ben Goertzel and Lisa Rein protecting – amongst many subjects – coaching AGI’s to carry one another accountable, pingable IDs (utilizing cryptographic hashes to safe agent identification), AGI rights & a lot, way more! (Sept 2024). And yeah, I’m there, too.
6. An article by Anthropic CEO Dario Amodei makes factors much like Reid Hoffman and Marc Andreeson, that the upsides of AI are doubtlessly spectacular, as additionally portrayed in a small however influential variety of science fiction tales. Alas, his listing of potential advantages, whereas in depth re methods AI may very well be “Machines of Loving Grace,”* can also be lengthy within the tooth and nearly hoary-clichéd. We have to recall that in any ecosystem – together with the brand new cyber one – entities with out suggestions constraints quickly evolve into no matter kind greatest proliferates shortly. That’s, until suggestions loops take form.
7. This article in FORTUNE makes a case much like my very own… that AIs will enhance greatest, in accuracy and sagacity and even knowledge, if accountability is utilized by AI upon different AIs. Optimistic suggestions could be a harmful cycle, whereas some sorts of unfavorable suggestions loops can result in incrementally elevated correlation with the actual world.
8. Once more, my featured WIRED article about this – Give Each AI a soul… or else.
My associated Newsweek op-ed (June’22) handled ’empathy bots” that feign sapience and personhood.
== AI generated visible lies – we will cope with this! ==
9. A brand new AI algorithm flags deepfakes with 98% accuracy — higher than another device on the market proper now. And it’s important that we maintain creating such techniques, so as to stand an opportunity of maintaining in an arms race in opposition to those that would foist on us lies, scams and misinformation…
… just about precisely as I described again in 1997, this reposted chapter from The Clear Society – “The Finish of Images as Proof.”
Two issues. First, scammers will use applications like this one to assist good their rip-off algorithms. Second, it might be silly to rely on anyone such system, or only a few. A range of extremely aggressive tattle-tale lie-denouncing techniques is the one factor that may work, as I focus on right here.
Oh, and third. It’s inherent – (as I present in that chapter of The Clear Society) – that lies are extra simply detected, denounced and incinerated in a basic atmosphere of transparency, whereby the best quantity can step away from their screens and algorithms and examine them to precise, physically-witnessed actuality.
For extra on this, right here’s my associated Newsweek op-ed (June’22) handled ’empathy bots’ that feign sapience. Plus a YouTube pod the place I give an expanded model.
== Generalizing to innovation, generally ==
10. Conventional approaches to innovation emphasize concepts and innovations, typically resulting in a shedding confrontation with the mysterious “Valley of Demise.” My colleague Peter Denning and his co-author Todd Lyons upend this paradigm in Navigating a Stressed Sea, providing eight mutually reinforcing practices that energy skillful navigation towards adoption, mobilizing individuals to attempt one thing new.
=== Some tacked-on tech miscellany ==
11. Certain, it goes again to neolithic “Venus collectible figurines” and Playboy and per-minute telephone consolation traces and Eliza – and the film “Her.” And bot factories are onerous at work. At my 2017 World of Watson keynote, I predicted persuasive ’empathy bots’ would arrive in 2022 (they did.) And shortly, Kremlin ‘honeypot-lure babes” ought to grow to be ineffective! As a result of this deep weak point of male people may have an outlet that is… greater than human?
May that result in these males calming down, prioritizing the extra necessary elements of life?)
12. And therefore, we are going to quickly see…AI girlfriends and companions. And this from Vox: Persons are falling in love with — and getting hooked on — AI voices.
13. Kewl! “This tiny 3D-printed Apple IIe is powered by a $2 microcontroller .” With a teensy working display taken from an Apple watch. Can run your previous IIe applications. Dimension of an previous mouse.
14. Paragraphica by Bjørn Karmann is a digital camera that has no lens, however as a substitute generates a textual content description of when & the place it’s, then generates a picture by way of a text-to-image mannequin.
15. Daisy is an AI cellphone utility that wastes scammers’ time in order that they don’t have time to focus on actual victims. Daisy has “informed annoyed scammers meandering tales of her household, talked at size about her ardour for knitting, and offered exasperated callers with false private data together with made-up financial institution particulars.”
And eventually…
== NOT straight AI… however for positive implications! ==
And… solely barely off-topic: In case you really feel a necessity for an inspiring story a few fashionable hero, mind and deeply-wise public determine, attempt Choose David Tatel’s VISION: A memoir and Blindness and Justice. I’ll vouch that he’s one of many wisest wise-guys I do know. “Imaginative and prescient is charming, smart, and fully partaking. This memoir of a decide of the nation’s second highest court docket, who has been with out sight for many years, goes down like a cool drink on a sizzling day.” —Scott Turow. https://www.davidtatel.com/
And Maynard Moore – of the Institute for the Examine of Faith within the Age of Science – will likely be holding a pertinent occasion on-line in mid January: “Human-Machine Fusion: Our Future, or May We Do Higher?” The IRAS webinar is free, however registration is required.
== a lagniappe (puns supposed) ==
Within the Nineteen Eighties this supposedly “AI”- generated sonnet emerged from the next immediate, “Buzz Off, Banana Nostril.”
Nicely… actual or not, right here’s my haiku response.
In lunar orchards
A pissed apiary indicators:
“Bee gone, Cyrano!”
Depend the layers, oh cybernetic padiwans! I’m not out of date… but.
