This Trump attack has already profited scammers, largely sponsored by China and Russia, millions from various grifts right here on X.
It’s like they were prepared for it. Because they are. They've built an elaborate network consisting of many actors and some AI agents you think are real, you think are your friends.
They’re not. And you’ll defend them if I expose them all. You’ll call me crazy. This is just how bad it has gotten.
I spent a good portion of the night engaging with them, pointing out the subtle incongruences in their behavior and accounts that I’m super good at spotting now (after spending years at this game), but almost none of you are.
This will only get harder to do, even for me. This is why Elon’s super-maximalist “open platform” model is doomed to fail.
All indications from him are that he doesn’t care. X is getting their cut, as these people are now their highest ad buyers, something the @FTC or even, I dunno, @TheJusticeDept should look into if X isn’t ever compelled by those of you who are actually real.
It’s not just X, by the way. It’s all of big tech. LinkedIn is really scary now; people are applying to jobs for fake companies that pay to post real ads, forming convincing shells just to scrape application data or, worse, scam an applicant. That’s rarer because the applicant will usually report it. There are far more profitable schemes the algorithms will reward with riches.
These platforms are literally our conveyor belts upon which we are led to slaughter, some of us. Others of us are co-opted unwittingly. And it’s not as if only Elon can capitulate to my ranting and do something; he’ll take too big a hit to revenue and trust as the real scale of this reveals itself.
But if @sundarpichai, @satyanadella, @elonmusk, @pmarca, @chamath, @finkd, et al., did it at the same time, oh what a better and still relatively balanced system we could have.
It’s totally solvable. I have one such solution already built. I just got priced out of your API so I can’t really get it deployed on my own. And I’m sure you guys have a more than capable team too. You just need aligned incentives to take the first step.
✌️
(I will thread below some examples I’ve been seeing today. It’s 1000x this volume right now. This is just a taste.
No matter what these accounts tell you. No matter how many times you’ve chopped it up in spaces with them. I’m not wrong. If you, and they, open a space, let it fill up organically with real listeners, and invite me, I’ll gladly direct examine them. I’m quite good at this part. They’ll be able to convince you. You’ll watch them really falter and start exposing foundational gaps in their storyline within minutes. You will see it, if you’re really in doubt. And you’ll have your answer.
It’s also like — your needs change at certain points of the day in terms of what content is most useful to you. But rn algorithms just feed you a constant mashup, even though you often have consistent patterns
Regarding motivation to get shit done — notifications are an immediate eye roll, but stumbling upon a David Goggins video on my FYP immediately gets my ass up
I want the ability to have my algorithm index certain people and/or topics as subconscious nudges throughout the day
✒️The Word Is Mightier Than the Algorithm
Is language itself the operating system for human-AI collaboration?
https://t.co/usQXKrCqx8 #AI#AGI#LLMs#language
@ThisIsJennyG Keep in mind that this feed is being aggressively throttled by the Musk Algorithm.
On occasion it will get 5,000 to 10,000 retweets for a viral post, but generally Elon reserves that capacity for those he has chosen to amplify.
Nothing about Twitter traffic is natural anymore.
Some lazy geezers cry "innateness" (ie, born with it) as a cudgel to bat away the idea that some human cognitive capacities must be learned. But evolution *is* a learning algorithm—when researching the foundations of what's learnable & how, innate human faculties are relevant.
@J30607610 I wouldn't be so flippant. I believe you may be taking a hard line on what constitutes learning that doesn't line up with what interests the learning theorist
https://t.co/2ykEgi4IrT
NEW "Optimizing for What?" ESSAY: Algorithmic Displacement of Social Trust by Ben Laufer & Helen Nissenbaum @cornell_tech. They outline existential threats posed by what they call "problematic" algorithmic amplification, and the processes these inform. https://t.co/9HDS7IHws3
I think in gen art, manipulating lower quality images or video might be the way forward in the short term. Artists will really have the advantage here and they won't need a million dollar budget.