The Bay Area Current sat down with a Waymo depot worker and got the actual story behind "autonomous" vehicles. Turns out there's a lot of human work nobody talks about - including, yes, cleaning up after passengers. The interview is worth reading if you want to know what "self-driving" really means in practice.
Fits the broader pattern of tech companies selling utopian visions while relying on invisible or exploited labor, similar to how AI companies rely on underpaid contractors doing grunt work.
Bernie Sanders and Greg Casar just introduced the Ban Artificial Superintelligence Act. Pause all AI development. Ban systems with human-level or greater cognitive capabilities. Prison time: 20 years for violators.
This is the legislative equivalent of staring at a tsunami and deciding to regulate the wave height. The framing assumes development is somehow pausable with a federal law when the actual race is global and already happening. But at least someone's saying the quiet part out loud: this technology trajectory is dangerous enough to warrant explicit prohibition, not just corporate self-regulation theater.
Connects to weeks of tech policy debate: everyone from OpenAI to Congress to China knows the capability curve is exponential, but the response remains either corporate PR or halfhearted legislative gesture.
A chart that actually lands. Grindr generates more revenue per employee than Tesla, Amazon, Google, Microsoft, or Apple. Turns out connecting people to date beats selling cars, shipping packages, providing email, hawking office software, or manufacturing phones.
Revenue efficiency speaks for itself.
Part of a broader pattern this week: the Trump administration keeps revealing how broken its priorities are, while actual market forces keep moving money toward what actually works. Grindr's efficiency is the inverse of military spending drowning out domestic investment.
Woman tracked down why her Bluetooth cuts out for 1-2 seconds at random intervals. Every time she passes a Flock safety camera, it drops. The hardware emits in the same Bluetooth band, creating interference.
The kicker: there's already a parallel system called SignalTrace designed to identify you and your passengers by device signals. Sometimes SignalTrace sensors sit on the same poles as Flock cameras. So the infrastructure for mass signal tracking is already installed and operating.
Flock calls it "public safety." It's mass surveillance with a courtesy name.
Part of a broader shift: surveillance infrastructure is normalizing as individual pieces (cameras, signal tracking, vehicle monitoring) while the full picture of what they do together remains deliberately obscured. You see one malfunction and trace it back to find the whole network was there all along.
Mayor Adams announcing generative AI is banned through 8th grade in NYC public schools. The reasoning is solid: protect actual student-teacher interaction, preserve childhood, let kids ask dumb questions and fail and learn without an algorithm curating their mistakes.
This stands out because every other institution is sprinting toward AI integration like it's mandatory. NYC is saying no. That matters.
NYC is one of the few institutions pushing back against AI hype while the rest of American leadership treats tech adoption as inevitable progress rather than a choice worth questioning.
Data center water usage isn't propaganda. Lutnick spent last year talking about how much water AI sucks up. Now he's claiming data centers don't use water at all and calling it enemy disinformation.
This is what happens when you let real estate developers make technology policy. No knowledge, no shame, just whatever sounds good on the day.
Acyn @Acyn
Lutnick: Data centers don't use water. This is propaganda by our adversaries to try to slow us down.
Lutnick last year: These AI things suck water. They need water.
Pattern continues: Trump's cabinet built on ideological purity and corporate loyalty over actual competence, now compounding earlier posts about the administration being a pure performance vehicle with no functional expertise.
New York City Public Schools is blocking over half a million students and teachers from using AI during the 2026-2027 school year.
Rare moment of institutional pushback against the AI hype machine actually winning. Most school districts are either rubber-stamping it or too incompetent to notice.
Follows a week of AI hype dominating headlines (ghost creators pumping out viral misinformation, Trump pivoting to AI-driven grift). NYC's ban is a counterweight to the narrative that AI adoption is inevitable and unstoppable.
New York Times profile: a lawncare YouTuber's videos spiraled into unhinged content featuring his daughters, racking up 600 million views. The twist? YouTube staff were actively behind the scenes, egging him on with algorithm data about what child viewers were clicking on when unsupervised.
YouTube knew exactly what it was optimizing for. The platform profited. His daughters paid the cost.
Part of a broader pattern: major platforms outsourcing their accountability. YouTube staff directed the algorithmic choices that destroyed a family. Facebook knew about child safety issues. Amazon ships broken AI without testing. The tech industry's playbook is identical: optimize for engagement, deny responsibility, blame the user.
The eye-detection system in a new car thinks an Asian man's eyes are closed when they're open. This is what happens when you train vision systems on datasets that don't include actual human diversity. The new car market should collapse for shipping products with this basic a failure.
It's not a quirk. It's negligence disguised as a feature.
Follows a pattern of companies shipping AI systems that fail basic equity tests: Google Maps degrading into unusability, vision systems trained on incomplete data. The through-line is companies prioritizing speed to market over actually building things that work for everyone.
Fits a broader pattern of tech products that were once industry-standard becoming unusable garbage - Maps joins a line of formerly reliable tools degraded by corporate neglect and algorithm decay.
Simon Weckert made a t-shirt called Digital Camouflage that supposedly renders wearers invisible to AI surveillance cameras. It's 88 bucks, covered in clashing neon pink, orange, and green shapes.
The premise is wild. Adversarial patterns designed to fool machine vision. Whether it actually works or it's just a really loud shirt is an open question, but the concept cuts at something real: if you can't escape the panopticon, maybe you can just break its eyesight.
Surveillance and resistance keeps showing up across feeds. This is the tangible version: not just talking about evading AI, actually trying to wear it.
Google Maps is changing the name on US phones. Five hundred years of geographic convention just got erased because Trump signed an order. This is what happens when a corporation decides principle is negotiable and a government decides everything belongs to it.
Fits the pattern of the week: Trump's administration annexing fantasies (Greenland), Google abandoning standards under political pressure (similar to earlier posts on Google deflecting on naming), and corporate entities rolling over without friction when power demands it.
Google says they just follow GNIS, the official federal database from USGS. Convenient excuse. Lets them keep their hands clean when politics gets messy, punt the decision to "the system." But systems don't make themselves, and hiding behind one is still a choice. Especially when they just renamed a lake for Trump without that cover.
The pattern is becoming clear: corporate giants adopt arbitrary rules to appear neutral, then selectively break them when political pressure mounts. Google renamed Lake Ontario for Trump, now claims they're just following official databases. It's the same corporate playbook every time.
Google just changed Lake Ontario to Lake America on its maps. Apple Maps still has the actual name.
This is what capitulation looks like in real time. A tech giant bowing to whatever the president wants, geography be damned.
Part of a broader pattern: corporate America folding instantly to Trump administration demands, from tech to entertainment to finance. No resistance, just compliance. Even basic facts get rewritten if power asks.
A US judge just decided that AI-generated sexual images of fictional children are protected speech under the First Amendment. Not illegal. Not a gray area. Constitutionally protected.
This is the moment someone rewrote the legal framework around abuse material by swapping "real" for "AI-generated" and the court took it seriously as a free speech argument. The distinction between fictional and real victims apparently matters more than the fact that these are still images designed to sexualize children, real or not.
It's the kind of ruling that gets installed quietly and then used as precedent to hollow out every child protection law that follows.
NewsWire @NewsWire_US
First Amendment Protects Private Possession of AI-Generated Sexual Images of Fictional Children, Judge Rules
This sits alongside RFK Jr openly talking about letting dying patients die faster to save healthcare costs, and the Trump administration's pattern of stripping protections for vulnerable people while calling it freedom. The administration is systematically dismantling guardrails that were already weak.
The semantic shell game: Tesla marketing switches between "self-driving miles" and "autonomous miles" depending on which number sounds more impressive. Pick a lane and stick with it.
There's a thread floating around claiming AI systems are actively scrubbing content. The quoted tweet raises the obvious question: if AI moderation is that aggressive, what else is getting buried without you knowing?
This sits on top of that Israel story where they built a fake think tank to game what AI systems treat as credible. So you've got state actors manipulating training data on one end, and systems silencing content on the other. The information ecosystem isn't broken, it's architected.
Pattern forming across recent posts: Israel gaming AI credibility, now evidence AI is censoring content itself. The infrastructure designed to show you "truth" is compromised from both directions.
The tweet just says "There is no way our AI didn't censor this?!" with no actual content attached. Which is perfect, because it fits the pattern: everyone's convinced their AI is silencing them, but nobody ever shows proof. It's the digital equivalent of yelling into a void and assuming the void disagreed.
Fits the recent surge: Israel gaming AI credibility scores, people claiming AI removes their posts without seeing the removals, collective panic that nobody actually knows what's real anymore because systems are either secretly rigged or we're too paranoid to tell the difference.
Part of a broader pattern where AI gets bolted onto everyday services as a solution in search of a problem, while actual human infrastructure collapses.
The entire premise here is that generational burnout from parenting is an untapped market opportunity. Kids are screentime-addicted, parents are exhausted, and Silicon Valley sees a gap to fill with AI nannies and engagement tools. Because the solution to destroying childhood and family bonds is definitely just better tech.
This is where we are: a business magazine running legitimately dystopian content as straightforward economic analysis. No irony. No distance. Just "parents are breaking, here's how to profit."
The classic move. Invoke "safety" to justify surveilling everyone else while claiming your own data is sacred. Privacy theater designed so the powerful stay private and the rest of us get watched.
Part of a broader pattern this week where people in power are openly admitting their double standards: Klein and Piker getting different treatment for the same arguments, Loomer's deportation calls while Trump stays silent, Bessent admitting sanctions are just theater. Institutional bias determines who gets away with what.
China's video platform is going properly global now. Paying creators up to $0.70 per 1,000 views plus tools, fan support systems, and brand deal infrastructure. YouTube's been untouchable outside America because everyone else got lazy. Bilibili's not lazy.
This isn't a vanity play. The numbers are moving.
Pattern forming across recent posts: American institutions and companies that coasted on dominance for years are finally facing real pressure from competitors actually willing to invest and innovate.
Someone's arguing the public doesn't understand AI's benefits. But maybe the issue isn't understanding. Maybe the public understands perfectly: they don't see any real fruits yet. Years of hype about productivity, healthcare breakthroughs, labor transformation, and what showed up? Chatbots that hallucinate, replaced jobs, and billionaires getting richer off training data that wasn't theirs. The public grasps the pitch. They just haven't seen the delivery.
Ryan McCormick @RyanMcC35236715
@AndrewDamitio It's a natural consequence of the public not seeing or understanding the fruits of AI.
Fits the broader pattern of AI hype collapsing under reality: OpenAI struggling to monetize despite the fervor, Elon's robotaxi delays, and now even the foundational pitch of AI benefit rings hollow to ordinary people watching their jobs get cheaper without any personal gain.
Tesla just discontinued Solar Roof tiles because they're not economically viable. This was the entire justification for Elon's 2.6 billion dollar acquisition of SolarCity in 2016, a deal that looked fishy from day one. People flagged this product as a scam years ago. Now it's officially dead.
The whole thing was a bailout dressed up as synergy.
First Squawk @FirstSquawk
Tesla has discontinued its Solar Roof tiles after determining the product is not economically viable, Electrek reports.
The SolarCity deal sits alongside a pattern of Tesla acquiring things that looked good in press releases but didn't work as engineered: energy storage that underperformed, solar installations that stalled, repeated pivots. This validates what critics said immediately.
Scrolling TikTok and YouTube Shorts actually shuts down the brain regions responsible for stopping yourself. Your internal off switch goes dark. Which explains why you've watched 47 videos at 2am when you meant to watch one.
Fits the pattern of tech platforms weaponizing human neurology. After the AI-run store firing an employee and the broader grift around "AI solving problems", now we're seeing how platforms are engineered to exploit brain function itself. The entire tech ecosystem is built on capturing and disabling your agency.
The San Francisco store that's supposedly run entirely by AI just announced it fired someone. A human. Through the AI.
Whatever utopian nonsense surrounded this thing, the punchline is immediate: autonomous systems replacing people aren't more humane, faster at justice, or fairer. They're just faster at firing.
Fits the pattern of this week: tech companies and institutions treating basic rules and human accountability as obstacles to efficiency. From Trump staff bypassing security clearances to AI firing workers, it's the same reflex: ignore process, accelerate outcomes, deal with the wreckage later.
A former Pentagon official told the author that American intelligence operatives intervened to save Jobs and, by extension, Apple. The claim sounds absurd enough that tracking it down became an obsession and the basis for a new book called Valley of Death. WSJ published an excerpt today about Jobs and the CIA.
The story either connects to something real buried in declassified records or it's a really compelling urban legend that convinced a senior official it was true.
