Theme: AI

  • ZOOM’s NEW ZOOM-IQ FEATURE (text summary) #Zoom’s new AI that records summaries

    ZOOM’s NEW ZOOM-IQ FEATURE
    (text summary)

    #Zoom’s new AI that records summaries of your meetings is ‘ok’. It’s not as good as the third party products whatsoever, but given that zoom is worried about Google’s introgression in the space, ZoomIQ is just a first step that needs work but that will improve.

    I’ve asked ChatGPT to summarize the ZoomIQ summary in under 5000 words (our meetings are content heavy), and the result is at least a fair list of the topics covered, even if the detail is overly general and not insightful.

    So for the time being we’ll stick with adding these summaries to our videos on YT.

    Note that our work is quite academic and dense, and impenetrable to the average audience, despite our use of humor and frequent failed attempts at humor, so we have dedicated but small audience of students who follow the work.

    I’m reluctant to share this Video content directly on XTwitter because while we hear placating words, the actions of the new CEO mimic the actions of the previous. And her most recent publications that suggest a return to previous suppression of uncomfortable truths, even when in neutral academic prose, and free of hostile intent.

    Those things that cause us conflict need the light of day so that we can resolve the conflicts and move on. I know Elon knows this but without legal protection from interference in marriage, employment, business, or services, it appears that the ‘suppression of resolution of conflicts by the suppression of discussion’ is very hard to reverse. Even though it’s prohibited under common law, until the age of social media, we didn’t have the legal need to restore that common law, to contemporary legislative code. (This is something our organization intends to pursue by class action and restore from common law to legislative code)

    So while we prefer the Twitter platform, now that we have long-form text, we are staying with Youtube and Odysee for videos where, as we are not seeking monetization, and only seeking to reach a dedicated adult, informed, and usually male audience, our work on the law and the so many taboos that need addressing in the law, that are otherwise impossible to discuss in social media without mouth-breathers seeking attention by offense seeking, find offense in even the science of the causes of our contemporary conflicts, and the mutually beneficial curese for those conflicts we have sought to produce for all. 😉

    Sorry. Had to stick that bit of truthful scarcastic wit in there for theraputic reasons. 😉

    Cheers

    Curt Doolittle
    The Natural Law Institute
    The Science of Cooperation


    Source date (UTC): 2023-09-24 19:43:31 UTC

    Original post: https://twitter.com/i/web/status/1706031628420521984

  • The argument we’ve been making (internal and external to MSFT) since the first u

    The argument we’ve been making (internal and external to MSFT) since the first use of video cards for calculating weights in the early 00s, is that if AI produces the search results we want, there is no place for advertising, and that kills Google’s revenue stream.

    Google’s problem is finding a value prop if advertising isn’t it. MSFT can float search without advertising indefinitely, as it has a revenue source embedded in an unbreakable network effect.

    And google had the first mover advantage and sat on it for years, precisely because they understood what we had almost two decades ago.

    There is an obvious counter to MSFT revenue stream for both Google and Apple. Apple is riding on the iPhone and Google on ad revenue. But these are both vulnerable.

    While I’ve never had a problem getting access to MSFT senior staff, or doing major deals with there, it’s been impossible to work with google, because they’re more arrogant (overconfident) than MSFT in the 90s.

    We change when we have to. And given how lame Claude is vs ChatGPT, and the direction they’re taking with it to avoid doing what they ought to, even if it’s a threat to search, the future ought to be interesting.

    Right now, it’s rather fun watching the industry try to brute force the LLMs into something other than a UI for a bottom up AI. And while progress is impressive, they have to solve the logic and reasoning issue ,world model issue and adversarial competition issue. And that’s before we get to needling neuromorphic hardware to end the memory and training compute problems.

    Cheers

    Reply addressees: @Scobleizer


    Source date (UTC): 2023-09-23 20:38:22 UTC

    Original post: https://twitter.com/i/web/status/1705683042612674560

    Replying to: https://twitter.com/i/web/status/1705678876058538362

  • Neologism: “Botserevations” (Bot Observations) ( I should headline my posts on o

    Neologism: “Botserevations” (Bot Observations)

    ( I should headline my posts on observations of normies in real life with the topical heading of (Botservation). 😉 )


    Source date (UTC): 2023-09-23 18:07:40 UTC

    Original post: https://twitter.com/i/web/status/1705645120740044941

  • Neologism: “Botserevations” (Bot Observations) ( I should headline my posts on o

    Neologism: “Botserevations” (Bot Observations)

    ( I should headline my posts on observations of normies in real life with the topical heading of (Botservation). 😉 )


    Source date (UTC): 2023-09-23 18:07:40 UTC

    Original post: https://twitter.com/i/web/status/1705645120689704960

  • —“Many people are already outperformed by LLMs on many intellectual tasks, but t

    —“Many people are already outperformed by LLMs on many intellectual tasks, but that’s in part because many people are habitually deepfaking general intelligence.”— Joscha Bach

    Ouch. Harsh. True.


    Source date (UTC): 2023-09-23 06:03:26 UTC

    Original post: https://twitter.com/i/web/status/1705462859788869861

  • Intelligence isn’t operationally complicated. It’s simple. Just a volume of very

    Intelligence isn’t operationally complicated. It’s simple. Just a volume of very expensive biology or hardware. “Enough Continuous and recursive memory” is all it takes. So Intelligence isn’t complicated per se, it just requires a lot of memory, a hierarchy, predictive auto association, valuation, attention, wayfinding, the ability to retain state of competing networks. Recursive wayfinding is just thinking, reasoning calculating and computing. All of those facilities are present in basic animals. The rest is just the scale made possible by the size of the brain and the number and density of neurons,

    LLMs are interesting in that they are brute forcing auto association from text as if it was experiential memory. It’s using text narratives instead of episodic memory. It has some vague attention-like facility. But it doesn’t have near the capacity for subsequent adversarial competition, and analysis and deconstruction facilitating falsification to produce logical reasoning. In this sense they are still probabilistic search engines. Though they can evolve the full suite with enough working memory, and can learn in real time with neuromorphic hardware.

    Reply addressees: @barbarikon @Plinz


    Source date (UTC): 2023-09-23 05:54:07 UTC

    Original post: https://twitter.com/i/web/status/1705460514027872256

    Replying to: https://twitter.com/i/web/status/1705378505289314578

  • RT @Plinz: I am unconvinced by the idea that “we” can recognize AGI when “we” ca

    RT @Plinz: I am unconvinced by the idea that “we” can recognize AGI when “we” can no longer distinguish it from our own performance. Many p…


    Source date (UTC): 2023-09-23 05:34:03 UTC

    Original post: https://twitter.com/i/web/status/1705455465310638291

  • Just in my feed from today alone: Medical Imaging Genetics Add mate selection Ad

    Just in my feed from today alone:

    Medical Imaging
    https://twitter.com/MissAndyAUS/status/1704702406490263974

    Genetics
    https://twitter.com/PaoloShirasi/status/1704540906886443126

    Add mate selection
    Add friendship Selection
    Add neighborhood Selection
    Add working together
    Add Speech Patterns
    Add Personality Differences
    Add (Big) Intelligence…

  • RT @MissAndyAUS: @PaoloShirasi AI can accurately predict race from medical imagi

    RT @MissAndyAUS: @PaoloShirasi AI can accurately predict race from medical imaging.

    Apparently our internal organs adapt to reflect socia…


    Source date (UTC): 2023-09-21 11:37:14 UTC

    Original post: https://twitter.com/i/web/status/1704822089344024807

  • It’s unfortunate because while GPT is aware of my work through 21 and has a vagu

    It’s unfortunate because while GPT is aware of my work through 21 and has a vague understanding of it, it would require training to be useful.

    We’re kind of frustrated that the only solution is to publish papers and populate web sites with the material necessary to train it,…


    Source date (UTC): 2023-09-21 03:18:38 UTC

    Original post: https://twitter.com/i/web/status/1704696609273974853

    Replying to: https://twitter.com/i/web/status/1704670289538552186