poe's law comes into full flower

and why the future may be local

meet providence assemblyperson “lyle culpepper.” please give they/them your attention as they speak on the recent shooting at brown university.

this is riveting stuff.

obviously, this presser would seem to raise any number of quite serious concerns, but i’d like to raise one that an awful lot of people online seem to be missing:

there is literally no parody of woke belief so absurd that it cannot be readily mistaken for reality if one lacks knowledge of a speaker’s priors (or even of their reality)

poe’s law has come into full flower.

there is no “lyle culpepper.” this is not even a person. it’s an AI construct, 1’s and 0’s playing make beleive with politics and humanity alike. there is no there there, just ghosts in the machine and gags on the reel.

the point i find most interesting here is not political, but rather psycho-social:

there is an impending reality fracture coming. the ability to tell what is real and what is fugazy is already well past slippery and into “you haven’t got a prayer without specialized tools.”

maybe you can say “oh, OK, but i can see it” but mostly, i don’t believe you. you see it because you look for it and your mind demands to find it. trying to pick the really high end AI out of a lineup of 9 real videos of similar content is rapidly becoming a sucker bet that very few people would take even (or even 2:1) odds on. (i wouldn’t)

at the speed people consume internet content, the likelihood of find the fake discernment drops to something indistinguishable from zero.

and it’s going to continue getting better at a speed that’s hard to conceive of. keep in mind that just a year ago, this was nowhere, way on the other side of the uncanny valley, obvious, silly, and rare.

no mas.

this is now cheap, ubiquitous, and astonishing.

is this a real singer with some special effects? or is this a full AI construct doing techno mongolian throat singing of christmas carols? is it a real face? a real voice? seriously, how would you know? do you even care?

this is going to upend industries in a fashion people are only just starting to grasp.

estimates are that 30-40% of new music being uploaded into streaming services is AI generated. some are trying to mark it as such. others are rapidly realizing that they basically cannot tell and that their detection tools are losing the arms race

advertising is in for a massive shakeup.

this video was made in one day using nothing but text prompts and a few hundred bucks worth of VEO credits. i’m pretty sure it was one writer.

if you had put this on TV, would anyone have noticed it was all fake?

we’re well into the realm of “there is no functional difference here” apart from 20 times the speed at 1/1000th the cost. and that’s going to keep dropping.

and as production costs approach zero, production quantity will soar towards infinity. keep that fact in mind, more on this is a minute.

people say “well, yeah, sure, whatever-whatever, i care not for the loss of hollywood jobs. heck, i applaud it! music too! they have not written anything good in decades. let the rains come and wash it all away!”

sure.

but what about the rest of it?

news.

internet videos.

internet videos that fool the news.

edits of edits of edits faking out reporters and viewers alike?

how can you trust anything you did not see with your own eyes?

was that trump speaking or was it an avatar?

to describe what’s coming as “impossibly borked” is charitable. this is going to be an unholy mess of indiscernible media, messaging, photography, and fakery.

this stuff is going to become deadly serious and “reality fade” into fractal digital fog will stress even the highest trust networks. tom stoppard (RIP) joked about “well how do i know there is such a place as denmark? it could be a conspiracy of cartographers.”

that particular piece of poignancy is about to become the existential riddle of everything.

the solipsistic slide into a non-verifiable external sphere is upon us.

maybe you can trust someone to tell you “what happened in shanghai today.”

but how can you know?

did this really happen?

one can conceive of some sort of “reality hash” where a camera has a private key and embeds a hashmark/watermark into the image such that you can verify that it is the unaltered output of that specific camera and then the images when posted can be second-signed by the poster’s private key to validate who put it up and when, but this necessitates a vast new technological system that i am unaware of anyone even working on at present and even if this did come to be, i’m not sure how much it helps.

you’d get some verification of key figures: politicians, leaders, CEO’s, known reporters etc. you can become a verified and known quantity.

but this is dangerous. it puts media back into the hands of the few, and the few lie. a lot. their agenda is not your agenda.

(it’s also a surveilance state’s wet dream.)

but even this is not really going to matter.

most of social media is NOT like that (and it’s not going away). it’s video from some person on the scene, someone you never met before who captured some event or videoed/photographed some going on. you have no idea if “bobjones67” from arkansas is real or made up of if their image hash comes from a nikon camera or just used the key they extracted from one (or if the picture is a real photo of an AI construct generated to look real in photos).

this is currently dated tech.

the time between now and when full “reality fracture” of inability to discern digital from meatbody is measured in months, not years.

the time to “i have no idea if the person on this zoom call is an avatar controlled by a person or an avatar controlled by an AI” is similarly demarcated. honestly, for high end users, they’re already there.

it’s just not widespread in the wild yet.

price drops, ease of use rises, integration surges so that you can use AI to write the graphics prompts to drive the graphics engine and pretty soon you’re really starting to ask some questions about what is real and what is not and the amount of “is not” gets so high that you start to question the knowability of reality altogether.

this way madness lies. people are neither willing nor able to live in these kinds of ubiquitous funhouse frauds. you just can’t. the psychological strain of constantly trying to parse a reality that slippery and facile is more than can be borne.

and the fraud and danger grow exponentially.

most businesses i know of now have layers of security around money transfer. you want to send a wire or an ACH? no one will do that based on an email anymore. in many cases, not even a 2 factor login. most institutions require voice confirm. you need to call the finance team from a known number and in a known voice confirm the details of the wire request. and even that fails.

i know a company whose CFO got fooled by an AI generated voice modulator pretending to be the CEO and calling from a faked caller ID that looked like the CEO’s phone.

they got a wire sent.

(fortunately, wires being wires, it was able to be recalled. this is one of the core problems with the no-trusted counterparty systems like bitcoin. reversibility is a feature people care a great deal about and “a couple days to settle” can be a feature rather than a bug.)

this is going on all over. frankly, i’m astonished it’s not happening more. how are there not fraudsters posting online videos of CEO’s saying things that make their stocks tank or double?

how are fake voices not placing orders? how are fake avatars not taking zoom calls? they proably are. honestly, they certainly are. the question is not “if?” it’s “how often?” and the answer can only increase logarithmically.

“i can soundz hooman?”

and we do NOT have anything like the infrastructure to defend against it either technologically or cognitively.

it seems like an awful lot of this could start to look like “lights out, game over.”

how on earth can you build far reaching and sufficiently flexible networks of trust in a system that could flip from being 80/20 real/fake to 99 fakes to every 1 real in very short order?

because that’s where “near zero cost, near infinite ubiquity” takes you. that it goes 2 orders of magnitude worse.

then the AI’s start learning you need to keep just enough truth in the mix to keep the players playing and start working you over with image linking and ordered impression framing and all manner of other cognitive trickery that is barely understood by humans but that rapidly emerges from large sample size A/B tests analyzed by agents with massive working memory so far past human scope as to be not only incomprehensible but quite possibly invisible.

to even touch such a substrate without AI mediation to protect you would be like shotgunning a can of cognitive ebola.

much blood, many fouls.

and so i have a niggling little pet theory here scratching around my little cat brain, one i cannot quite shake.

it may sound a little weird, but bear with me and keep an open mind. sit with this a bit before you render too firm a judgement.

i think this may be a very good thing.

i think it may be an excellent thing, a profound thing that can reshape humanity and make it much, much better, happier, and more flourishing.

hear me out.

this all sounds like a disaster, but that’s because you’re thinking about it from inside the frame.

if you basically lose the ability to trust anything, you cannot see, any text or information about the world, any photo, any video, disembodied voice, zoom face or phone call, any of it, what do you do?

you stop trusting it.

the world gets small again. you trust what you see with your own eyes. business moves to highly protected, verified networks, but even these are not really trusted because the arms race to break any new safeguard will always be ferocious and AI mediated fraud and intrusion is a level of relentless attack on every surface at once all the time every time that will make some bigtime penetrations inevitable.

so for the real high trust? face to face. IRL. hands get shaken, eyes get looked into, and people once more do business directly with people.

i think this has the potential to make people happier.

we’re social animals. we want to be around one another. we want face time not facetime. we want contact, lunch, talking and body language not the half communication of voice and and video calls and the quarter communication of text.

the internet is too big a room for humans to thrive in. it’s too vast, there are too many comparisons, too much information, too many lies and shapings and framings in the always on A/B tested demolition derby for slivers of attention span and the hotwiring of limbic responses.

it’s too important to walk away from. you could not compete with the others who stayed. they would be too much faster, know too much more. but if it all devolves to hot, predatory garbage assualting your and your perceptions in wave upon wave of AI slop, lies, and outright cognitive assault and mindrape that starts to look more and more like evoked hypnosis and the cultivation of mass delusions and the madness of crowds for fun, profit, and political purpose, maybe everyone walks way.

it’s hard to remember life before the internet. “jesus, how in hell did we even do that? how did we manage?” those in gens Y and Z perhaps have no memory of life without mobile data, social media, all of it. it was slower, that much is for sure.

but it was also happier. humans were not as anxious, as wound up, as inflamed into pigpiles wars of all against all, tribe against tribe, and of minds evolved to hunt deer and eat berries in caves and sit around fires telling stories to one another against life in a room too big for human mechanisms to calibrate properly.

what if the information age devolving into the age of utter lies and untrust collapsed distance?

what if you stopped caring about what was happening on the other side of the world because it once more became vague, untrusted, and basically unreal?

what if it reversed the zoom diaspora and brought us back to city centers? what if it brought us back to towns and town centers? what if we left the house to meet people and find dates and mates instead of scrolling and swiping?

might this restore quite a lot of humanity to the humans?

might it take back so much of what i think so many of us can feel has been lost? replace the loss and alienation that seems endemic, especially among the digital native generations like Z? might the deracinated once more plant roots in meatspace and realize that the meaningful parts of lives are only forged IRL?

if you ask most people in this purportedly modern age if we have lost the path and lost the plot, it seems like they would tell you yes. we can argue about degree, but about the fact of that matter? not so much.

many joke about “the matrix was correct, 1999 was the apogee of human civilization.”

well, that was right before the internet changed everything, wasn’t it?

and maybe going back is not such a bad idea.

i’ll tell you one thing:

if this is even half correct, then the pennies on the dollar sale prices of class A office space in urban centers are one of the best buys in a generation.

and the cities that can get their acts together and become safe, fun, flourishing, and hospitable, open for business, open to business, and ready to let the creators and builders run free, the ones who sort out their schools and stop tolerating the homeless industrial complex of petty crime and ostentatious mental illness and drug addiction, those cities will win the land grab for talent and become the new centers of the world.

it feels to me like a once in a century, maybe once in a society kind of jump ball may be going up into the air here.

will be fascinating to see who has the prescience and the presence of mind to come down with it.