An Everything App to support your whole life. Will you want it?
PLUS a new product to guess, some extra helpings of recent stories, the shamelessness of another high tech founder dude, and Meta’s cool Ray-Bans
Shout out to 2 readers who sent me links to articles. One with more gloomy news of AI hiding stuff from us and the other about a completely AI DJ on an Australian radio station. For my young readers radio is an outdated 100+ year old technology that shoots sounds waves through the air to be captured by a special receiver to play that sound back to you. Like very primitive Spotify. Only more fun.
EXPERIMENTAL RECIPES
MAIN COURSE: Big Tech to support your entire life with the Everything App, which is surprisingly small
Silly Con valley has been driven by killer apps for decades now. The first one was the spreadsheet VisiCalc which kicked off the PC revolution in 1979. (For my younger readers that was not the political correctness revolution; that came a few years later). Big tech has chasing after these holy grails ever since. So now both Elon and Sam - first name techbros, though Altman may not yet be quite there with seniors over 75 - are going to blow this quest wide open with the great super app that does absolutely everything you need.
I think I’m dreaming. You mean just one lonely icon on my iPhone? One solitary but muscular app running on my laptop. Cross platform too. That syncs to my phone, iPad, desktop and Smart TV? Does everything from soup recipes to selecting lug nuts?
Of course frigging’ not! This is the worst of Silly Con valley meeting our technology fuelled hyperbolic era where a sneeze is an epidemic and every third person is a fascist. (By the way the sneeze joke was not some anti-Covid coded remark because we live in the all-conspiracy epoch as well).
Musk’s vision “is to combine social networking, messaging, payments, streaming, e-commerce, and even banking into one seamless digital platform” He also said you won’t need a bank account. He has recently added AI (read Grok the X-based AI tool) into this conception. It has been his stated objective for X for some time.
Notice anything about this vision. We are basically passive slaves, talking to “friends” in isolation while we spend money on delivered goods and stream entertainment services. Real things like walking in nature, meeting friends, and working seem to be missing. The goal is to turn IRL (in real life) into an anachronism while autonomously driven trucks deliver goods made by robots ordered by one of your many digital agents while you talk to AI bots.
This is such a spiritual and enriching outlook for humanity that it makes me glad I’m 70.
Altman’s vision is less compelling which is why he doesn’t yet qualify for the single Sam sobriquet. His is of a social network tightly tied to crypto and blockchain, mainly to prove you aren’t AI. You can also purchase stuff using mini-apps tied to Worldcoin because spending is all we consumers are supposed to be doing. Just speculating here but his emphasis on proving you are real must be driven by some powerful nightmares.
Pragmatically, even though I know millions will, I’m not trusting my hard earned money to these….guys.
EXTRA HELPINGS: Bad rhymes and cliches update the cheating story
We have “truly” passed the Zuckian “move fast and break things”, to you snooze you literally lose in the speed of light age. Last week I wrote about Cluely helping candidates cheat on video interviews . The bits had hardly settled in my Substack post when a product named Truely was announced to detect people using Cluely. If you started hearing unusual sounds in your workspace, it is because this newly released software cruelly triggers an alarm when it sniffs out the use of Cluely. It worked as the cheating marketing line was duly changed to helping with sales calls (where evidently we expect massive lying and cheating).
Cooly, the kids …er founders behind Cluely announced a hardware initiative to get around Truely’s software based solution. With typical bravado, ““Whether it’s smart glasses, a transparent glass screen overlay, a recording necklace, or even a brain chip, we’re not sure,” he said.” We might soon see them again as an upcoming Experimental Recipe. A reader recently asked where I get these products. From people like this.
QUICKBYTES: AI model builders are getting desperate for more training data
There you are, trending super cool in your new Ray-Ban Meta glasses as you stroll down the boulevard to pick up your morning 6 adjective iced coffee drink. You stop to have a few words with a former colleague, smile at a passing women and linger at a shop window eyeballing the newest Samsung Neo QLED 8K TVs. Unless you have gone to great lengths to opt out of your data being scooped up for AI training purposes (some features like voice don’t have an opt out), you can never be sure that what you said, saw or looked at hasn’t been recorded and used to train AI models ( and prompt you to buy the TV you looked at or join a dating app). Sure they say that they aren’t recording voice until you say Hey Meta and that photos / videos are under your control. But this is Meta, do you trust them not to turn the rest of your life into data to use, package and sell?
They also look a little clunky and heavy to me.
EXTRA HELPINGS: Robust study on AI model accuracy
For those people wondering - as I am - about the massive amount of work in business organizations that is probably filled with errors based on the study of people lying about using AI and then not checking the results, comes a little rigour about using AI models. Are we making erroneous decisions or changing strategies based on crucial misinformation provided to us?
We don’t know the answer yet but are increasingly realizing how precarious facts and truth are. At a AI research workshop with 31 participants assessing one model in substantial depth came the two critical observations:
“LLMs can product vastly different output even with the exact same prompt and data”
“The supporting quotes the LLM provided looked okay superficially, but on closer investigation *every single participant* found examples of data being misquoted or entirely fabricated”
I have strongly suspected the former as I will often as the same question later and get a somewhat different answer, and I’m doing trivial search-like prompts. Even though the women running the research workshop cautioned that this was one prompt, with one dataset, and one model, I think that our use of AI in business organizations needs to be thought through thoroughly. Oops that may not work as I forgot about the liars.
QUICKBYTES: The Instagram glasshouse throwing stones at Generative AI
In one of the most hypocritical statements that I have recently read - that went unchallenged in the media ecosystem we have evolved - is a co-founder of Instagram, Kevin Systrom claiming that AI models “juice engagement” by asking follow-up questions. So clarifying what you might really want, a common practice in any human interaction, is considered juicing. This is breathtakingly brash. Instagram is one of the most maligned social platforms whose algorithmic design means your feed doesn’t show all whom you follow and promotes unhealthy new accounts, their deliberate focus on disappearing stories over regular posts, and constantly suggesting tantalizing reels and AI chatbots must make them world-wide juicing royalty. I personally wouldn’t open my mouth given the many studies showing mental health issues Instagram caused in teenage girls, as but one example. Judging by his testimony at the current Meta trial I guess Systrom is laying all these issues at Zuck’s feet, a Silly Con version of blaming the current economy on Biden.
MENU MISTAKES
I realize Canada is a tiny country ( size-wise would be a huge 51st state so what would Texans and Alaskans do) compared to the USA but I’m not sure these items would make any of our TV networks as accomplishments. Plus we got rid of our penny yonks ago.
Another Tonic with some hopefully tasty takes. So many crazy tech stories, so little time. Thanks to my readers. Comments, submissions, messages, likes and restacks all greatly appreciated. We need a dialogue as we gradually sink on the Technology Titanic. Grab a deck chair.
So basically, we're heading toward a future where Musk wants to be our wallet, therapist, best friend, bank, baby daddy, entertainment system, and probably our AI pastor, all rolled into one app? 😂
Can’t wait for the future where I have to blink twice to prove I’m not a bot before buying overpriced eggs?
Silly Con - I like it. More people need to use it.
Happy Tuesday David.
I spent enough time in the Unix world to appreciate the value of lots of small well-tuned apps that do specific things. So, the idea of a "one app to rule them all (and in the darkness bind them)" doesn't appeal to me. At all.
In the 1950s, Isaac Asimov wrote "The Caves of Steel" and "The Naked Sun" in which the "spacers" (the segment of humanity that went to the stars) live in individual isolation surrounded by robots and technology and only interact with other humans via holographic transmissions. Actual contact with others makes them squeamish. We seem headed in that direction. Hence, perhaps, the recent meme about "touching grass".
Zuck looks like Napoleon Dynamite.
We're seeing more and more that Ai confidently delivers wrong answers it thinks the users want to hear. I'm starting to wonder when it'll be responsible for some major disaster.
Is there an error in that FauxNews graphic? Do they mean *plastic* straws? (Or is that the menu mistake? I'm not entirely clear on the concept.)