Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror

Comment Ignore this troll. (was Re:It doesn't) (Score 2) 142

No, it really doesn’t. And neither does your post. What it does is hijack a discussion about the loss of a vital climate data resource—critical to public safety and long-term economic planning—and derail it into a jaded screed about how “everyone is too immature to save democracy.” It’s a logic fire sale: false dilemmas, bad-faith assumptions, historical revisionism, and a parade of straw men so dense they’d block NOAA radar. You hijacked a vital discussion about the loss of scientific transparency in disaster data and climate response, and replaced it with lazy cynicism, bad history, and a cascade of logical fallacies. The only thing bipartisan here is the contempt you show for every side of the debate—while contributing nothing but noise.

Trump can basically do anything he wants and no one can stop him because the only serious threat to his power is that the Democrats would win in a landslide taking a supermajority in the Senate.

Fallacy: False Dilemma.
You’re setting up a binary where either Democrats win a supermajority or Trump becomes omnipotent. That’s not how power works in a system with multiple levers—courts, states, civil society, investigative journalism, and yes, voting. This statement is defeatism masquerading as analysis.

In order for that to happen the Democrats would have to wield the power they have in order to shut down voter suppression. This would involve basically ignoring court orders as needed...

Fallacy: Advocating Lawlessness as Strategy.
You just argued that Trump is a threat because he breaks democratic norms. Your solution is for the opposition to also break democratic norms? That’s not fighting fire with fire—that’s burning the Constitution to own the fascists.

They absolutely have the power to do it but the problem is the kind of centrist Democrat who would be in charge... does not under any circumstances want to wield power.

Fallacy: Straw Man.
You’ve built a caricature of centrists as ritual-obsessed cowards. It’s convenient for your rant, but it ignores real-world cases where they’ve pushed legislation, defended voting rights, or reformed procedures—just not always with the reckless abandon you demand.

Now ordinarily what would happen is the left wing of the party and the left wing Independence would give them a kick in the rear...

Fallacy: Nostalgia-Based Reasoning.
There’s no historical basis for your fantasy scenario. When exactly did the far-left successfully prod moderates into coordinated voter-rights action in defiance of court orders? Oh, right—never.

The problem with that is the left wing doesn't really want to stop voter suppression...

Fallacy: Motive Fallacy + Projection.
Claiming the left “doesn’t want” to stop voter suppression is not just bad faith—it’s maliciously dishonest. There are entire orgs, movements, and lawsuits dedicated to this effort, many led by those very progressives you sneer at.

They have a childish dream of kids 18 to 24 showing up to vote in droves...

Fallacy: Ad Hominem.
Mocking younger voters as delusional ignores data showing that Gen Z turnout is rising and increasingly decisive. Is it idealistic to want better policy from younger, more climate-conscious voters? Sure. But idealism isn’t immaturity. Dismissing it as such is cynical rot.

...they have literally been trying to achieve that for so long that the 18 to 24 set they originally started working on are in their 60s and voted overwhelmingly for Trump...

Fallacy: Historical Revisionism.
Boomers were never “targeted by the left” in their 20s to enact Scandinavian welfare states. That’s a fever dream with no citation. Boomers have been a slim majority in his column, not a monolithic voting bloc. In 2016 and 2020, voters aged 65+ gave Trump 52%; in 2024, that slipped to 51%. That’s not overwhelming—that’s barely treading water in an increasingly aging demographic pool, and that support is actually decreasing. Misrepresenting that as generational fanaticism isn’t just lazy, it’s you lying with a straight face, period.

Basically we are a nation of 12-year-olds. We act like children. Left right up down center doesn't matter...

Fallacy: Hasty Generalization.
This rhetorical nuke is supposed to sound profound, but it collapses under its own weight. If everyone is childish, then no one can be reasoned with—which conveniently excuses you from offering actual solutions.

It's possible the sheer incompetence of Donald Trump and everyone around him might save us again...

Logical inconsistency.
You just claimed Trump can “do whatever he wants,” but now you're hoping his own ineptitude might doom his plans? Pick a narrative and stick to it. Or better yet—engage with the article topic you’re hijacking.

I suspect the whole country is going to turn into a fascist dictatorship like China and or Russia...

Fallacy: Slippery Slope + Fear-Mongering.
You’ve now left the realm of argument and entered a nihilistic prophecy. This is political clickbait disguised as despair. It’s emotionally manipulative, analytically hollow, and contextually inappropriate given the thread topic.

I hope my kid can flee the country when it happens and that Europe has enough nukes... I'm also hoping to die before the worst of it.

Emotional Blackmail.
This level of performative doomcasting is designed to shut down debate, not invite it. And it does nothing to advance understanding of why cutting a disaster database during an era of record-breaking billion-dollar weather events is outrageously shortsighted.

Would love to be proven wrong but I have watched these last several months...

Fallacy: Anecdotal Fallacy.
Your political despair, while real, is not evidence. And it does not justify hijacking a thread about weather disaster data loss to rehash your disillusionment with the Democratic Party.

Every now and then one of the lefties realizes how fuck they are and does a screed... they're hobbyists in it for the fun so that's not going to happen.

Fallacy: Poisoning the Well.
You don’t get to accuse others of being unserious while writing an off-topic rant with zero citations, no actionable suggestions, and maximal contempt for everyone involved.

Comment Re:baseload? (Score 1) 56

Your entire argument is a false binary dressed in engineering drag. It's a textbook case of energy punditry by omission: the math checks out only because the premises are rigged. It cherry-picks capacity factors, ignores cost curves, handwaves away deployment timelines, and treats energy policy like a cage match between lone contenders instead of a collaborative grid. What you are spinning is a pro-nuclear sermon disguised as analysis—one that forgets solar is already winning where it matters most: in economics, speed, and scale. If you're going to build your argument on selective stats and outdated assumptions, don't be surprised when someone shows up with the rest of the data—and a longer memory. Go shill somewhere else.

Imagine a 5+ GW nuclear power plant like that at Barakah.

Sure. Barakah is real, and ~5.6 GW when all four reactors are online. No argument there.

There are four reactors at the plant, each reactor needs some down time for refuel and maintenance, so assume at any time there's a reactor shutdown which means it is producing 75% of rated output 24/7.

And here’s the trick. That 75% is a convenient understatement. Global average capacity factors for nuclear are closer to 80–90%. U.S. reactors often exceed 92%. Assuming 75% sounds cautious, but it subtly tilts the table.

The capacity factor is likely better than that but for the moment assume 75%.

This is the rhetorical confession: acknowledge the lowball, then build the rest of your case on it anyway. It's a magician's misdirection—tell just enough truth to hide the hand.

On the other hand you have solar power, which has a capacity factor of about 20%.

In cloudy regions, yes. In sunny ones—like the UAE, or the American Southwest—utility-scale solar reaches 25–28%. You quote selectively to make solar look worse than it is.

...solar output isn't 100% but for a few hours every day at best, since clouds could rain on that parade.

This is not analysis, it’s doomcasting. Clouds are real. So are grid-balancing strategies. Forecasting, geographic spread, time-of-use pricing—none of these are new ideas. We don’t cancel aviation because of turbulence. We don’t cancel solar because of clouds.

If people want to push back on that then I can assume 25% capacity factor, if only to make the math easier.

Another rhetorical sleight-of-hand: concede the better number, but only after you’ve primed the reader with fear, uncertainty, and doubt.

So, to make the same energy over a year as a 5 GW nuclear power plant with a capacity factor of 75% there would have to be 15 GW of solar power with a capacity factor of 25%.

This is mathematically correct—and economically misleading. Solar doesn’t need the same cooling water, safety buffers, or siting restrictions. “Overbuilding” is part of the design strategy and priced in. You’ve done the math, but skipped the context.

For solar power we'd need to see batteries on the grid...

Correct—but not unique to solar. Batteries stabilize the grid. They help nuclear too, which doesn’t load-follow well. Storage isn't a solar tax. It's a grid-level upgrade.

On the other hand we could use nuclear power that provides power at night so we aren't using batteries for that.

False dichotomy. Nighttime demand doesn’t mandate nuclear. Wind often peaks at night. Batteries don’t care what charges them. And nuclear’s inflexibility becomes a liability, not a virtue, when demand ramps up and down.

...Because a 5 GW nuclear power plant is providing that power day or night the need for batteries is reduced...

Only if your grid demand is perfectly flat. It isn’t. So unless you’re willing to dump excess power or pay consumers to use it, you’ll need storage or curtailment. Either way, the economics wobble.

I do not believe batteries will save solar from competition with hydro, geothermal, nuclear fission, or even wind.

And now we pivot from numbers to belief. But belief doesn’t stop the reality that solar + storage is already cheaper than new-build nuclear in almost every market on Earth. Not in theory. In contracts.

Onshore wind is so cheap that it would be stupid to not build that out to the largest capacity possible.

Agreed. So is solar. And they’re complementary. Wind is stronger at night and in winter. Solar dominates midday and summer. This isn't a zero-sum game. Real grids use both.

Geothermal and nuclear fission apparently work best (as in cheapest) when run at a steady state...

True. But demand isn’t steady. And these plants take a decade or more to build, with massive up-front costs and political risks. That’s why they're losing ground to flexible, modular solutions.

Hydroelectric dams are awesome for low cost, reliable, low CO2, and safe energy...

Also true. And also geographically limited. Climate change is making hydro less predictable in many regions. Droughts and reduced snowpack are already cutting capacity.

Where’s the value add in solar + batteries?

You mean aside from the fastest deployment timeline, zero fuel cost, modular scalability, no cooling water, and the ability to build on rooftops, deserts, and brownfields? Aside from the fact that solar + batteries is the only combo scaling fast enough to matter in this decade?

We’d need massive overbuild... and massive batteries...

Yes. Just like we already overbuild gas to handle summer peaks. Just like we built massive coal fleets to handle winter demand. Overbuilding is how you build reliability—regardless of fuel.

Of course no electrical grid will rely on any one energy source...

Then why spend twenty paragraphs pretending we must choose between solar or nuclear?

My belief is that if we get low cost grid-scale energy storage then that will not be kind to the economics of solar power.

This is where the mask slips. If storage is cheap, solar wins. Every time. Because once built, solar has no marginal cost. Nuclear does. Every watt-hour stored and discharged costs less if it started life as a sunbeam rather than a uranium pellet.

Nuclear power will likely see the most benefit...

Only if you assume storage is cheap, but nuclear somehow gets it all to itself. That’s not an economic model. That’s a fantasy. Nuclear plus storage is like building a castle and then buying an armored caravan to haul water to the moat. Solar plus storage is a well in your backyard.

Comment The game to lower wages (Score 1) 125

This would be like requiring LAW as required course for k-12.
So everyone is skilled in Legal and now there is an over supply of lawyers. That causes lower wages.
So if everyone knows at least some level of programming then there will be more programmers so supply demand in the work force works in the company favor.

Comment GOP and the Partisan Dismissal of Reality (Score 2) 211

Looks like abandoning science wasn’t just a MAGA sideshow — it’s been structurally baked into GOP policymaking for decades.

Northwestern just dropped a brutal empirical study — Partisan Disparities in the Use of Science in Policy (Furnas et al., 2024) — showing that Republican policymakers systematically cite less science, cite worse science, and trust scientists about as much as they trust independent election auditors.

The findings are brutal: Democratic committees are nearly twice as likely to cite science as Republican ones, and left-leaning think tanks are five times more likely to ground their arguments in scientific research. Republicans, when they cite science at all, lean heavily on outdated, low-impact, or pre-peer-review work — the intellectual equivalent of presenting a stack of hotel bar napkins as expert testimony. Republicans aren’t cherry-picking from the same tree — they’re harvesting from a rotting orchard, and they’ve been doing it for a quarter-century.

Why?

Because the modern GOP — hijacked first by the Dixiecrats looking for safe harbor during the Civil Rights movement, and then two generations later by MAGA's mob of white Christian Nationalists and authoritarian nutbars — is a coalition built on resentment, reaction, and rejection of any authority outside its own tribal mythos. Of course they distrust scientists. Scientists have an inconvenient habit of insisting that gravity works, vaccines save lives, and the Earth is older than 6,000 years. But the GOP's vision of conservatism, especially the MAGA-infested version it is currently suffering under, demands more.

Frank Wilhoit's ruthless summary of conservatism nails it:

Conservatism consists of exactly one proposition, to wit:
There must be in-groups whom the law protects but does not bind, alongside out-groups whom the law binds but does not protect.

The MAGA conservatives running the current GOP want to advance and protect their privilege, not any conservative principle. Every scientific fact that challenges their privileged position in the power structure becomes the enemy. Science, like justice, demands impartiality. Conservatism, corrupted by MAGA into tribalism, demands exceptions.

You don't need a PhD in political science to draw a straight line from the southern "Dixiecrat" realignment during the Civil Rights movement to today's Project 2025 fever dreams. It's the same anti-intellectualism, the same theocratic longings, just duct-taped onto an internet-age grift machine.

And the kicker?

This paper robustly shows that even when both parties write policy documents about the same issue — say, minimum wage — they still cite completely different scientific studies. They're not building on a common foundation; they're arguing from different planets.

Which brings us to now. The GOP’s Project 2025 outlines plans to dismantle the administrative state — a media-friendly sound bite and polite euphemism for purging government scientists, researchers, and regulators who dare to apply empirical evidence to public policy. MAGA and their fellow travelers don't want science interfering with their agenda, period. Remember Dick Cheney’s infamous comment about “creating their own reality”? He wasn’t posturing. For once in his life, he was telling the truth.

So here's the question for my reasonable conservative friends, who also claim to be principled Republicans:

If you keep voting for a party that systematically rejects science, and has been doing so for decade after decade, what exactly did you expect would happen?

When you strap yourself to a cult of grievance and gaslighting, don’t look shocked when the tracks end, the ground falls away, and the laws of physics finish the argument. You bought the ticket. Enjoy the ride.

Comment WotC and CC: roll for litgation check (Score 1) 35

WotC has released its core D&D rules under a Creative Commons license. After all, if you're going to win IP battles in court, why not start with a quick victory in the court of public opinion?

The D&D faithful are celebrating—but given WotC’s litigious history, and a decade-long pattern of corporate shell games and IP hoarding—you’d think fans might be more skeptical. It’s a PR feint—genuflecting toward open source ideals while quietly setting the stage for their next round of IP-consolidating, license-restricting greed.

WotC’s PR history is a string of natural 1s. This is the gaming industry—IP is the only resource worth farming, and WotC knows how to till the wallets of gamers with surgical precision. I’m probably not the only one who remembers when some beloved IP assets—once under WotC’s grubby stewardship—were exiled into legal oblivion. Not because of innovation, but because another bloated gaming company with more lawyers than ideas decided to sue for copyright infringement.

WotC’s embrace of Creative Commons smells less like redemption and more like illusion magic—slick, shiny, and just convincing enough to lure you into range of the next monetization ambush. Don’t worry, they promise it only hurts if you fail your saving throw.

Comment Re:In other news (Score 1) 181

How many people have died drinking themselves to death? Number of pot deaths is still at zero.

Here’s the thing: F = ma doesn’t care if you’re stoned or drunk. When your deceleration event involves a bridge abutment, your bloodstream chemistry becomes academic. Sure, THC toxicity is negligible -- it is less destructive to your liver than alcohol, and it’s true that cannabis doesn’t have a direct overdose death toll in the same league as alcohol. But metabolic harm isn't the only issue when it comes to THC. You seem to be deliberately ignoring the systemic consequences of THC use.

Comment Verizon’s CEO for Consumers Trolls Net Neutr (Score 2) 76

“I still don’t know what problem we’re trying to solve.”

That’s Verizon’s consumer CEO, Sowmyanarayan Sampath, playing dumb in an interview with The Verge. The guy responsible for 115 million wireless and 10 million broadband lines in the U.S. doesn’t understand net neutrality—or pretends not to. Either way, it’s disqualifying.

Let that sink in. The executive who oversees a third of the nation’s connectivity admits he doesn’t grok the principle that your carrier shouldn’t get to decide what apps, services, or content load faster—or at all—based on what makes them money. And yet, he’s fully empowered to throttle, shape, or “manage” your traffic however he wants.

His justification for throttling you into oblivion is the same tired line tier-one ISPs have been peddling since the late '90s—back when sharp-eyed sysadmins were busting them with packet captures from Ethereal, the open-source sniffer that grew up to become Wireshark. It’s always the same story: blame the mythical bandwidth hoarder, hand-wave about “network management,” and pray nobody remembers when Comcast got caught spoofing TCP RST packets to silently kill BitTorrent connections—then flat-out denied it until the packet logs made lying impossible. I’m honestly surprised Sampath didn’t defend it as a necessary safeguard “against that one person in your town who’s sucking up all the bandwidth”—his exact words, mind you.

Because nothing says “public utility” quite like throwing an entire neighborhood under the bus to throttle a protocol you don’t like.

And just when you think the playbook couldn’t get any dustier, Sampath reaches for the most threadbare metaphor in the telco FUD arsenal: the firefighter in traffic. That’s right—packet shaping isn’t about corporate deals, it’s about saving lives.

        “If there are 300 people using a congested road, and if a firefighter needs to go through, people are going to move aside and let him go through. That’s all we are saying.”

Except that firefighter isn’t a first responder—it’s Disney+, or whichever content partner Verizon is bundling into your “unlimited” plan. This isn’t emergency access; it’s product placement. Click-through rates. Inflated subscriber counts. Sampath even brags that Verizon’s “non-connectivity business” is now a $15 billion operation—built on bundling, streaming tie-ins, and convenience-fee masquerades. And all of it rides on the very network he insists is neutral and fair.

This is where the sleight-of-hand becomes stage magic: throttle independent traffic, subsidize partner traffic, and declare it a win for consumers because technically, you’re paying less for Netflix. Never mind that you’re paying more for everything else.

And let’s not pretend this is new. Verizon has a long history of trying to own both the pipe and the content. Remember when Verizon acquired Yahoo? When they acquired AOL? When they launched Go90 and tried to monetize landscape mode? None of it worked, but the ambition never went away. Now, the strategy is subtler—bundle services like Disney+, Hulu, ESPN+, Max, and YouTube Premium, and quietly rebuild the walled garden. Verizon controls the gate, the gatekeeper, and increasingly, the content behind the gate.

And if you’re wondering how seriously Verizon takes the “rules of the land,” that Sampath is making a show of genuflecting towards, here’s the cleanest contrast you’ll see outside a packet trace. When the FCC tried to enshrine net neutrality under Title II—rules that literally said don’t block, don’t throttle, don’t extort—Verizon went to war. Lawsuits. Paid astroturfers. A full-court regulatory press. But today? The same FCC—reshaped by Trump appointee Brendan Carr, who literally wrote the Project 2025 chapter on turning the FCC into a partisan mouthpiece—is now threatening to block mergers over corporate DEI policies. And Sampath? He shrugs and says, “We have to follow the rules of the land.” Amazing how that phrase suddenly acquires moral weight—when the regulation leans the right way.

The next time Sampath assures you it’s all just reasonable traffic management, remember: this is a man who looked directly into a camera and claimed not to understand net neutrality. Whether it’s willful ignorance or finely-tuned dishonesty doesn’t matter—the result is the same. You’re not listening to a network engineer. You’re listening to a corporate tool whose one job is to sell you the congestion that his packet shaping policies create.

Comment Weaponized AI Isn’t Coming. It’s Here. (Score 1) 61

We talk a lot about autonomous weapons and drone swarms when we talk about weaponized AI—but that’s only one prong. The quieter, more insidious front is already here: surveillance capitalism, behavioral targeting, and large-scale manipulation campaigns driven by generative AI. The tech is mature enough, the incentives are misaligned enough, and the regulatory landscape is impotent enough that this kind of abuse isn’t a hypothetical. It’s infrastructure.

And in the U.S., it’s going to be especially dangerous. Why? Because fascist-aligned movements like MAGA have already built robust epistemic cocoons—sealed narrative environments that actively repel contradictory information. A weaponized LLM doesn’t need to out-argue anyone. It just needs to affirm what the user already wants to believe, in confident prose, with links. That’s not persuasion. That’s reinforcement. And it works disturbingly well when the target audience has already been softened up by decades of epistemic isolation, courtesy of Fox News, Infowars, and podcast sewer pipes like Joe Rogan.

Some of us saw this coming.

Back when the internet was still mostly plaintext and hope, we fought hard to preserve signal hygiene. I'm dating myself, but I was a USENET admin, and I was there during the Canter and Siegel spam wave on USENET, and I remember when someone unleashed the first proto-vigilante automation to stop it: the Cancel Moose. It was clever. It was principled. It worked—for a while. But we all saw what happened next: AOL opened up a monetized choke point to the internet for 6 million middle class American morons. The internet died that day, and the arms race between idealists and advertisers started. And it has tilted hard toward the monetizers ever since. Social media finished what the early commercial web started: turning attention into currency and disinformation into a business model.

The article gets this mostly right—LLMs are vulnerable not because they’re dumb, but because the entire ecosystem is now structured to reward manipulation. And we’re seeing nation-states unambiguously exploit that. But the real issue? It’ll be domestic actors, armed with ideology and venture capital, who do the most lasting damage. This isn’t the beginning of a war. It’s just the latest escalation in a long-running arms race—one that started the moment Sir Tim gave us markup for the internet, and venture capitalists decided they would define the rules of engagement.

Comment Re:It don't need to be russia or china or.. (Score 1) 61

There is a large mass of people that hate being used to train an AI without their consent, and these people can (and rightfully so) spike the data to purposefully ruin the AI in retaliation.
To not mention things like "tar pits" made to stick the scraping robots and all that.

You're not wrong. I’ve written before that I have no moral qualms about salting the well for any company that profits by converting human attention into model weights without consent or compensation. If you're training your LLM on the detritus of digital life, don’t be surprised if some of it turns out to be metaphorical broken glass and rusty nails.

Tar pits, data spike traps, poisoned honey pots—all of that is fair game when the scrapers show up uninvited. But here’s the rub: guerrilla warfare is a tactic, not a strategy. It can punish, delay, frustrate—but it can’t build. Not by itself. If all we do is jam their sensors, they'll just retrain on cleaner data next time. The asymmetry in compute and curation capacity still favors the large players. Always has.

I’m old enough to remember the first wave of spam and snake oil that tried to colonize USENET. We fought it off with cancelbots, killfiles, and an almost religious commitment to good signal hygiene. And we won some battles. But not the war.

So yeah, spike away. But let’s also be thinking about how we build countermodels, open datasets, and non-captured spaces where the value of the collective isn’t funneled upward and monetized behind our backs. Guerrilla ops are satisfying—but strategy is what leaves something standing afterward.

Comment Re:imagine (Score 1) 61

Your post is a mix of pseudo-insight and veiled antagonism wrapped in a rhetorical fog machine. It’s cynical, performatively exasperated, and vaguely anti-elitist, hitting all the usual notes: disbelief in progress, disdain for experts, and the obligatory swipe at Elon Musk to keep the posture edgy. What it doesn’t offer is coherence. Or accuracy. Or intellectual honesty.

You open with a false dichotomy—that unless AI can independently divine truth, it isn’t real AI. Then you drift into a cloud of cynicism as camouflage, implying all efforts are corrupt or self-serving without making a single concrete claim. You lean on a straw man, pretending that AI researchers just shrug and let models ingest misinformation. And you close with a category error, equating model training with moral child-rearing, as if transformers needed bedtime stories and a conscience.

Imagine an intelligence that could determine truth rather than having to rely on being told what is true, imagine an intelligence with some sort of ability to reason, perhaps a "reasoning" AI.

This reads like you are frustrated that AI doesn’t conform to your sci-fi expectations. We have plenty of systems that perform bounded reasoning—symbolic solvers, theorem provers, decision engines—but LLMs aren’t designed to “determine truth” in the metaphysical sense. They model language, not reality. The idea that a system should magically infer truth from a sea of contradictory human output without any scaffolding is like being angry that a microscope won’t toast your bagel. The tools do what they're designed to do.

Funny how words get thrown around, yet we all know they are all lies.

This isn’t analysis. It’s performance. Accusing everyone of lying while remaining intentionally vague is a classic bait tactic—it invites the reader to feel superior without having to prove anything. It’s also functionally useless unless you’re writing copy for a MAGA bumper sticker.

We won't have actual AI as long as it is vulnerable to this kind of pathetic manipulation, but we all know the power and the money is in this very manipulation so don't hold your breath on that problem being solved.

Here’s that false dichotomy: either AI is completely invulnerable to manipulation, or it’s fake. By that logic, the human brain is also not “actual intelligence,” since people fall for scams, cults, and clickbait daily. Every intelligence system is vulnerable to manipulation—including us. The point isn’t to eliminate manipulation entirely but to recognize and mitigate it, which is exactly what the article (and ongoing research) is about.

Also, don't we all know to teach our children right from wrong? Yet somehow we don't know to teach AI the same?

Your category error here has Gilbert Ryle spinning in his grave. LLMs aren’t moral agents. They don’t have values, experiences, or a theory of mind. They don’t learn lessons. You don’t “teach” them in the same way you raise a child—you select, weight, and filter input data, fine-tune model behavior, and set boundaries via reinforcement or rules. The idea that this is some kind of moral failure is nonsense dressed up as insight. When—and if—we ever build a true AGI, your lament might start to make sense. But for now, you're literally barking up the wrong tree.

We accept NOT curating the information the AI's train on, or worse yet making sure they get trained on misinformation and lies? Just how intelligent are these smartest people in the world?

Nice straw man. No, researchers don’t “accept” training on lies. Curation is the entire point of modern alignment and model integrity work. The issue isn’t that no one is trying—it’s that attackers are now gaming the system with deliberate noise injection. That’s not a design flaw of AI, it’s a consequence of making it open to the same polluted infosphere we all swim in. Pretending otherwise is like blaming the firewall for malware that was emailed to you by your cousin.

Comment Re:How does that work? (Score 1) 34

How can a judge rule something to be unconstitutional but at the same time permit a violation of the constitution? That makes absolutely no sense. Either it's wrong and has always been wrong and is never permitted or what was done in this case was legal and remains legal because it does not violate constitutional rights. Regardless of how you feel about this, the ruling makes no sense for that reason.

Because the warrant violated the constitutional rights of other people but not the defendant. The warrant should have been issued to search only for the defendant's phone, and that would have been constitutional. Had the tower dump been done without a warrant, then the data would have to be tossed. The others besides the defendant could possibly sue for Fourth Amendment violations, but not the defendant.

You’re actually circling the right idea here—unlike the GP, who’s flailing around like he just discovered the Constitution was written in English.

The key thing to understand is that constitutional violations and evidentiary suppression are not the same thing. The judge in this case didn’t say “the search was fine.” She said the search was unconstitutional, but that the evidence wouldn’t be suppressed because the officers relied on a warrant issued under precedent that hadn’t yet been overturned. That’s the good faith exception in action (see: U.S. v. Leon, 1984), and it’s been settled law for decades.

You’re a little off when you say it “didn’t violate the rights of the defendant.” It absolutely did—the court found the warrant to be a general warrant, which is forbidden under the Fourth Amendment. The issue isn’t who was harmed so much as how the evidence was gathered, and whether that method violated constitutional protections.

And the idea that the warrant would’ve been constitutional if it just targeted the defendant’s phone? That’s exactly what the ruling implies. Narrow, specific, based on probable cause? That flies. A mass dump of thousands of unrelated users’ data? That’s a Fourth Amendment foul.

So credit where due—you’re sniffing out the right path here. So, a little clarity...we need to untangle a few things: whether a search was constitutional in the first place, whether any evidence obtained should be suppressed because of it, and whether the law was clearly established at the time the search occurred. Those are separate legal questions, and conflating them is where most of the confusion—and troll-fuel—comes from.

It’s not a logic fail—it’s a balancing act. Constitutional law is a system designed to protect rights without detonating every case the moment precedent shifts. Anyone demanding binary outcomes is either trolling, or just really confused about how law actually works -- or both, in the case of the GP.

Comment Re:Make policing hard (Score 1) 34

I hear you. Truly. I used to say the same thing, sometimes louder and with fewer commas. You’re not wrong to be skeptical of expanded law enforcement powers, especially in the post-9/11 climate. But your post takes a legitimate concern and wraps it in a kind of performative absolutism that undercuts the very scrutiny you're calling for.

Ever since 9/11 we have done everything possible to make life easier for law enforcement up and down the board at the cost of all of our rights.

I remember that mood vividly. The Patriot Act, NSLs, mass surveillance—some of it was genuinely alarming. But not everything has moved in one direction. Courts have pushed back. Carpenter v. United States was a major Fourth Amendment win. Even this Nevada tower dump case is proof that judges are willing to draw hard lines, even if they stop short of suppressing evidence. Painting the last two decades as a one-way erosion of rights might feel cathartic, but it also erases the actual legal wins that people fought tooth and nail to achieve.

Sure, requesting phone data is part of policing. But those requests often are either non-existent or get rubber stamped by the courts.

That’s half-true—and it’s the half that gets repeated until people stop checking. Yes, some warrants get rubber-stamped. But when those warrants are challenged in court—as they were here—they’re being scrutinized more than ever. You’re looking at a ruling that explicitly says: “This was a general warrant. This violated the Fourth Amendment. This is a problem.” If that’s not oversight, I don’t know what is.

We have situations where a bad guy gets away, the police cry about how hard it is to catch them because they don't have this or that, and municipalities and legislatures just bend over backwards to give them what they want.

Sometimes, yes. But treating every resource request or legislative proposal as a Faustian bargain isn’t analysis—it’s fatigue in disguise. The answer isn’t to reflexively oppose everything law enforcement wants. It’s to apply exactly the kind of legal precision this ruling represents: support constitutional tools, reject unconstitutional ones. If the courts had ruled the tower dump fine without a warrant, you’d have a point. But they didn’t. That matters.

We should be scrutinizing and questioning every single move a LEO makes in order to prevent the type of situation we're in now.

Scrutinize, yes. Always. But let’s not confuse vigilance with hostility. “Every single move” is an impossible standard, and frankly, a recipe for burnout and disengagement. What we need is targeted, well-informed scrutiny that draws real lines. This case is that. The court said: this warrant was unconstitutional. This kind of dragnet won’t fly in the Ninth Circuit anymore. That’s a win for the Fourth Amendment. No drama needed.

I get where you're coming from. I really do. But outrage isn’t a strategy—it’s a posture. We’re not in high school civics anymore. We’re in the domain of real law, real courts, and real progress that moves in fits and starts. If you want to make policing harder in all the right ways, this ruling is exactly the kind of pushback you should be applauding—not dismissing just because it didn’t overturn the system overnight or come with headlines and a slam-dunk ending.

Comment Bad Warrant + Good Faith = Real 4th Amend Problem (Score 1) 34

A federal judge in Nevada ruled that "tower dumps"—mass cell phone data grabs by law enforcement—are unconstitutional under the Fourth Amendment. But because the cops had a warrant and the law was unsettled at the time, the court didn’t suppress the evidence. Cue outrage, confusion, and several users rediscovering the concept of judicial discretion.

This article is from 404media, and to their credit, they’re actually covering a meaningful privacy case. But true to form, they can’t resist dramatic framing. The line “cops could, this one time, still use the evidence they obtained through this unconstitutional search” brushes up against clickbait—it’s legally accurate but designed to provoke confusion. Then there’s this gem: “the cops got a warrant to conduct the tower dump but argued it wasn’t technically a ‘search.’” That’s not just sensationalism, it’s explosively misleading. The *DOJ* argued that, not beat cops in a precinct lobby. If you’re going to report on constitutional law, precision matters.

That said, the ruling itself is a step in the right direction. The court explicitly recognized that tower dumps are searches, and that general warrants like the one used here are unconstitutional. That’s a win for digital privacy. The only reason the evidence wasn’t thrown out is because of the longstanding good faith exception, which prevents evidence suppression when police follow existing precedent and obtain a judge’s warrant. If that bugs you, good—it means you’re paying attention. But that’s a policy issue, not judicial hypocrisy.

So yes, the evidence stays—for now. But the court has drawn a clear constitutional line, and that matters. The next warrant like this in the Ninth Circuit? It’s going to have to be a lot more specific. That’s progress, not capitulation.

Slashdot Top Deals

Ernest asks Frank how long he has been working for the company. "Ever since they threatened to fire me."

Working...