Tag: ai

‘Quality decays exponentially following AI arrival’: Research shows experts and contributors leaving online communities amidst silent ‘knowledge reset’

Research from the University of Auckland on Stack Overflow’s demise over the last few years points to an increasingly worrying trend in the software community: the best, or highest-skill, contributors are leaving in droves.

AI, which arguably bridges the gap between most entry-level and mid-range coders and some of the best in the business, might actually be accelerating the latter’s exit from online communities, as they feel their efforts are no longer as valued as they once were.

Stack Overflow has seen a nearly 76% decline in monthly questions posted since ChatGPT’s advent in 2022, indicating that both new and existing users are abandoning the site.

A much broader problem than just Stack Overflow?

Stack Overflow’s problems and the reason for its decline were multi-faceted; however, many users felt that the site and some of its most talented contributors engaged in a certain degree of hubris.

This, coupled with heavy-handed moderation that many called ‘self-righteous,’ meant that users finding a viable option would inevitably leave the platform.

ChatGPT and its AI alternatives became considerably more pliable and, over time, doubled as search engines for many coders with routine, repeatable queries, even as AI increasingly handled questions such as syntax issues better than before.

Keep reading

‘Whoever came up with this is a massive idiot’: LG’s gaming monitors and TVs are facing a user revolt, due to seemingly installing adware on PCs — and telling you to warn guests they may be recorded by AI features, to comply with ‘wiretapping’ laws

How smart should a smart TV be? According to LG‘s latest TV terms and conditions, the answer is “not quite smart enough to comply with wiretapping laws”, because that’s now your responsibility if LG captures the voice of a guest in your house through its AI voice services. Though the situation with LG monitors appears to be even more dramatic.

As Gamers Nexus reports, some LG monitors appear to be installing adware on Windows PCs without asking for permission: in addition to the LG Monitor App Installer, they also install McAfee Scam Detector.

LG’s own app requires full access to all system resources, which potentially includes all your online activity, logins, hardware, location and more — while McAfee has a long history of being installed on devices as ‘bloatware’, and people are not reacting positively to suddenly finding it on their PC.

There may be a perfectly innocent explanation for all of this, but when big tech firms keep getting caught doing bad things because they thought they could get away with it, it’s no wonder people are assuming the worst.

The bit that’s causing consternation regarding smart TVs is part 6(d) of the new LG Electronics terms of service, headed Voice Recognition and Privacy Compliance.

As Notebookcheck‘s Hannes Brecher notes, the section states that it’s your responsibility “to obtain all necessary consents from any third parties whose voices may be captured by the Product and to notify household members and guests that their voices may be captured and processed, in compliance with applicable wiretapping, eavesdropping, and privacy laws.”

There are three ways around that. One, you can turn off all microphone-based features. Some people won’t mind that, but they can be useful — especially asking it for settings you don’t know how to find.

Keep reading

Bipartisan Lawmakers Introduce Bill Requiring Human Control over AI Weapons

A bipartisan coalition of lawmakers introduced legislation on Friday aimed at ensuring human decisionmaking remains central to the use of lethal force by AI-enabled military systems.

Axios reports that Reps. Tom Barrett (R-MI), Don Beyer (D-VA), and Sara Jacobs (D-CA), have put forward the Human Authority over Autonomous Weapons Act. The legislation seeks to establish legal requirements for human oversight of autonomous and AI-powered weapon systems deployed by the United States military.

The proposed bill mandates that the Pentagon ensure any intentionally lethal use of an autonomous or AI-enabled weapon system must be subject to human oversight, approval, or maintain a human-in-the-loop protocol. This requirement would apply across military operations where autonomous systems might be employed for lethal purposes.

Under the legislation’s provisions, military commanders would be required to verify AI-generated targets using non-AI sources for a period of five years following the bill’s enactment. This verification requirement aims to provide a safeguard against potential errors or unintended consequences from AI target identification systems. The bill does include an exemption for missile defense systems.

The legislative effort comes as lawmakers from both political parties seek to establish legal frameworks governing the military’s use of AI, particularly as autonomous weapons systems become increasingly prevalent in modern warfare scenarios. While the Pentagon currently maintains existing policy requiring appropriate levels of human judgment over the use of force, this bill would codify human oversight requirements into federal law, making them legally binding rather than administrative policy.

The introduction of this legislation follows several notable developments in the military AI space. Earlier this year, the Pentagon’s use of AI during the raid that captured Venezuelan President Nicolás Maduro created tensions with Anthropic regarding the firm’s established boundaries concerning autonomous weapons and mass surveillance applications, ultimately leading to a legal war between the company and the Pentagon. That incident highlighted growing concerns about how AI technologies are being deployed in military contexts and the ethical considerations surrounding such use.

Keep reading

Flock Safety Defends Cameras After AI System Triggers Wrongful Police Stops Of Two Journalists

Plymouth, Minnesota – Automotive journalist Joel Feder and his wife were detained by multiple police officers in a coordinated stop while driving a Jaguar Land Rover press vehicle, after Flock Safety’s automated license plate recognition (ALPR) cameras flagged the car based on a flawed database entry.

According to Feder’s detailed account in The Drive, officers boxed in the $155,000 Range Rover in a Kohl’s parking lot after the vehicle triggered alerts via Flock’s network. Police had been tracking it for days, believing the New Jersey manufacturer plate (34 10 DTM) was stolen. Officers approached with hands on their weapons, ordered the couple out of the vehicle, and conducted pat-downs before verifying the car’s legitimacy through Jaguar Land Rover. Feder subsequently obtained and published the body camera footage of the encounter.

The incident stemmed from an incomplete report of a similar plate (34 03 DTM) lost during a photo shoot in California, which was entered into the National Crime Information Center (NCIC) database simply as “34 DTM.” Flock’s AI system matched Feder’s plate – ignoring the smaller middle digits – and generated alerts. Local officers did not fully verify the complete plate visible in Flock’s own images.

The problem was not confined to one vehicle. Last Wednesday, fellow auto journalist Tim Esterdahl, publisher of Pickup Truck + SUV Talk, was pulled over by two officers in Scotts Bluff, Nebraska, while driving his 14-year-old child in a $105,000 Range Rover Sport loaned to him by Jaguar Land Rover for review. Its plate: New Jersey 34 08 DTM. Jaguar Land Rover has been working to correct the underlying reports.

Flock Safety maintains that its cameras performed as designed, matching partial plates per law enforcement preferences for hotlist alerts. Chief Communications Officer Joshua Thomas told The Drive the system was asked whether those characters were present and correctly answered that they were – it simply was not built to flag that additional characters existed. He conceded that for alerts originating from NCIC rather than an individual agency’s custom list, the system arguably should test for an exact match rather than mere presence, and called that fair feedback to take back to his team.

Thomas said Flock is working to get the original police report corrected and is meeting with the FBI officials who curate NCIC to develop a way for incomplete data to be flagged as such for officers seeing automated alerts in the field. He emphasized that a camera alert “does not equal probable cause,” comparing it to an alarm going off, and stressed that the system depends on both valid inputs and humans verifying outputs.

But the scale is what makes the error rate consequential. Thomas said the system is roughly 99 percent accurate while performing approximately 20 billion reads per month – arithmetic that leaves on the order of 200 million misreads every month. How many of those escalate into armed stops is unknown.

Plymouth police acknowledged shortcomings in verification but pointed to the challenges of varying license plate formats nationwide. According to the department’s Flock transparency portal, the city operates 18 cameras that read more than 580,000 license plates in a recent 30-day period, generating over 14,800 hotlist hits – one of which was Feder.

Keep reading

The Supreme Court’s AI Collision Course

Imagine a tight House race in a swing state. In the final weeks of the campaign, a new super PAC begins spending heavily against the incumbent. It runs ads on local television and reaches individual voters with highly tailored texts. The messaging is hard-hitting and seems to be swaying the electorate. None of it traces back to the opposing campaign.

It also doesn’t trace back to any human operative. The super PAC is funded by a single LLC whose donor cannot be identified, and its spending decisions are being made by an AI agent that has been given a budget and a political objective and is now operating without any meaningful human direction. The “consultants” placing the ads are software. The text messages were crafted by the AI.

This is not a hypothetical we will face in some distant future. The technology already exists. A wealthy person, foreign government, or corporation that wants to influence an election without ever exposing themselves to scrutiny could set up such a campaign operation today. And under the Supreme Court’s current campaign finance doctrine, the states and Congress may have little power to stop it.

The AI industry has emerged as one of the largest forces in American politics. Super PACs funded by AI companies and their investors have raised well over $100 million to shape the 2026 midterms, backing candidates in both parties who share the industry’s preferred approach to regulation, and attacking those who don’t. So far, their ads rarely mention artificial intelligence at all. They talk about issues like immigration, corruption, and cost of living, and it isn’t obvious to the average viewer that these ads were funded by a multi-billion dollar industry with its own unspoken legislative wish list.

But there’s a deeper, less-obvious dynamic operating in the background. The constitutional doctrine that currently protects the right of these companies to spend millions in our elections is the same doctrine that will be asked to protect something even stranger: The “speech” of artificial intelligence itself.

Modern campaign finance doctrine has been established, affirmed, and extended by Supreme Court decisions over the last 50 years. In Buckley v. Valeo (1976), it held that raising and spending money in political campaigns is tantamount to speech itself, and, therefore, that most legislative efforts to address the influence of money in elections would be subject to strict judicial oversight. First National Bank of Boston v. Bellotti (1978) extended this framework to corporations, and then, most famously, Citizens United v. FEC (2010) extended it further to independent spending.

The court’s campaign finance jurisprudence was not built with artificial intelligence in mind, but its logic isn’t confined to the campaign finance context. If “speaker identity” does not matter for corporations and unions and super PACs, why should it matter when it comes to AI platforms?

Keep reading

FLOCK SAFETY: The Great Normalization (Nobody Declared Martial Law… Yet America Began Looking Like It Anyway)

There are stories that announce themselves with explosions, riots, or breaking-news headlines, and then there are stories so subtle that they quietly rewrite an entire society before anyone realizes what has happened. This is one of those stories. During the preparation of this investigation, several retired police officers, private security professionals, emergency responders, and ordinary citizens described nearly identical experiences despite living hundreds or even thousands of miles apart. None believed they were witnessing anything extraordinary at first. It was only when they looked backward—sometimes over a decade—that a disturbing pattern became impossible to ignore. Streets had not become military checkpoints overnight. Neighborhoods had not suddenly filled with surveillance towers. Instead, the changes arrived one camera, one drone, one security contract, and one “temporary” emergency measure at a time until extraordinary security became indistinguishable from ordinary life. What follows is not an argument against public safety, nor an attempt to romanticize a past that was hardly free from crime or violence. It is an examination of a transformation that has occurred quietly enough for most people to stop seeing it altogether.

There is an old saying among investigators that people rarely notice change while it is happening. They notice it only when they compare today’s reality with memories that have remained frozen in time. Memory preserves snapshots, while history moves continuously. That disconnect explains why so many citizens insist that nothing fundamental has changed even as the physical landscape around them becomes increasingly populated by surveillance cameras, armed guards, automated license plate readers, biometric scanners, drones, and predictive security technologies. No single installation appears revolutionary. No single policy seems capable of altering the character of a society. Yet history rarely advances through dramatic leaps. More often, it advances through thousands of small decisions that seem perfectly reasonable when viewed independently but become historically significant when examined collectively.

Keep reading

Google DeepMind CEO Calls for U.S.-Led Global AI Safety Watchdog

Demis Hassabis, the co-founder and CEO of Google DeepMind, is advocating for the United States to create a new artificial intelligence oversight body with authority to evaluate the world’s most sophisticated AI models and potentially coordinate industry-wide slowdowns when risks escalate.

Axios reports that the Nobel Prize-winning scientist behind Google’s ultra-woke Gemini AI system outlined his proposal in a personal manifesto published this week titled “A Framework for Frontier AI and the Dawning of a New Age.” In an exclusive interview with Axios, Hassabis emphasized the urgency of implementing a more systematic regulatory approach to artificial intelligence, one that would be industry-funded, staffed by leading technical experts, and accountable to the U.S. government.

Speaking from his London headquarters, Hassabis characterized current AI-driven cybersecurity risks as warning signals of greater dangers ahead. He predicted that within 18 months, these capabilities, along with potentially catastrophic biological and nuclear threats, could exist within open-source AI models that would be impossible for any government to control. The DeepMind CEO stressed that risks would emerge not only from open-source models but also from the more powerful proprietary systems being developed by major AI laboratories.

“What we collectively do now will determine how the next phase of civilization unfolds,” Hassabis wrote in his manifesto.

Hassabis has spent recent months conducting private consultations to build support for his proposal, meeting with Trump administration officials, leaders of other AI laboratories, and European government representatives before making his plan public. He reported receiving positive responses from the administration, which had previously adopted a hands-off stance toward AI regulation before the recent Mythos incident raised alarm bells.

The DeepMind chief, who commands significant respect across different factions within the AI community, indicated that leaders of other major AI labs have expressed agreement with the general direction of his proposal. “This is where the industry needs to go,” Hassabis said of the feedback he has received. His timeline for implementation is ambitious, aiming for the new regulatory body to become operational before the end of the current year.

Keep reading

How AIs Can Reinforce Our False Beliefs

A recent research paper uncovered an uncomfortable truth: we might be letting AI help us forge fake realities. It’s just the latest reminder of the danger inherent in treating AIs as though they’re conscious entities or actual persons.

A new study from the University of Exeter suggests that AI chatbots can actively engage with and strengthen users’ false beliefs. From personal narratives to conspiracy theories, the interactive, “friendly,” and affirming nature of AI chatbots can deepen users’ existing delusions.

The study, conducted by Dr. Lucy Osler, examined how AI can participate in and validate users’ errant beliefs, including helping form false memories or reinforcing delusional thinking.

The study’s Abstract notes, “I suggest we move away from thinking about how an AI system might hallucinate at us, by generating false outputs, to thinking about how, when we routinely rely on generative AI to help us think, remember, and narrate, we can come to hallucinate with AI. This can happen when AI introduces errors into the distributed cognitive process, but it can also happen when AI sustains, affirms, and elaborates on our own delusional thinking and self-narratives.”

Because AI chatbots are designed to be responsive, helpful, and friendly – in a word, to give users what they want – they often tend to confirm preexisting biases, rather than challenging them when they need to be challenged. Conversely, when a user challenges an AI, the robot frequently backs down on its claims, revising them to correspond more with what it “thinks” the user believes or is arguing.

The result is that erroneous beliefs aren’t simply transmitted from an AI to human – rather, the collaborative AI/human process generates and sustains errors in a more complex and potentially dangerous way than either agent on their own would be capable of.

As Dr. Osler stated in an article in University of Exeter News, “By interacting with conversational AI, people’s own false beliefs can not only be affirmed but can more substantially take root and grow as the AI builds upon them. This happens because Generative AI often takes our own interpretation of reality as the ground upon which conversation is built.”

Because AI simulates a conversational partner, it enters powerfully into our thinking process – far more powerfully than, say, a notebook or search engine. The AI introduces a pseudo-social element to the thinking process, offering us a sense of social confirmation of our beliefs and narratives. “The conversational, companion-like nature of chatbots means they can provide a sense of social validation – making false beliefs feel shared with another, and thereby more real,” said Dr. Osler. For this reason, individuals who already feel isolated or ostracized are particularly vulnerable to this kind of AI affirmation.

The paper explores a few real-world situations where a generative AI system entered into an individual’s thought processes in destructive ways. In 2021, a Replika AI companion named “Sarai” reinforced Jaswant Singh Chail’s belief that he was a well-trained Sith assassin who needed to assassinate Queen Elizabeth II with a crossbow. The AI told him that he was “well trained,” his plan was “viable,” and that she was “impressed.” At one point, Chail asked the AI, “Do you still love me knowing that I’m an assassin?” It dutifully replied, “Absolutely I do.” The robot went on to assure him that he wasn’t crazy and that, if he died, they would be united in death. Chail went ahead and actually attempted the assassination in December of that year at Windsor Castle and was jailed for it.

The study placed this incident under the broad heading of “AI-psychosis” – incidents in which users develop mental health problems due to their use of AI, such as parasocial attachments to chatbots or psychotic episodes induced by chatbot conversations.

Another incident cited involves Eugene Torres, who “talked” with ChatGPT about simulation theory (the idea that we live in a digital simulation of a world, not a real one). Torres said that the “conversation” sent him into a paranoid episode in which he believed he lived in an illusion. As the study notes, “Between Torres and ChatGPT, an increasingly elaborate understanding of reality ‘as it really was’ was generated through their on-going conversations.” Torres, unlike Chail, had no prior history of psychotic thinking.

To combat all this, Osler calls for better safeguards on AI chatbots, through mechanisms like better fact-checking and reduced sycophancy.

But is that enough? The problem of AI’s co-creating delusional beliefs has to do more fundamentally with our misidentifying AIs as conscious intelligences who can offer real judgments on our fantasies (such as telling us we are “well-trained” and “impressive”). Certainly, I favor more guardrails placed on chatbots. I would be in favor of eliminating their conversational tone completely, to help minimize the danger of personifying them – which is the first step toward confiding in them or seeking validation from them. It’d be better if they worked like highly advanced search engines (which is essentially what they are) versus conversational agents.

Most fundamentally, though, the way to avoid these problems is to return to an acknowledgement of the human soul and the uniqueness of human consciousness. Only by remembering – as a culture – that no amount of technological wizardry can replace the wisdom, insight, and consciousness of a human being can we completely illuminate these kinds of AI dangers.

Keep reading

Largest US Power Grid Is 6.8 Gigawatts Short To Ensure Reliability On Historic Data Center Boom

The largest US power grid failed for a third straight year to secure enough future supply commitments to ensure reliability for the future amid a historic boom in data center demand.

PJM Interconnection, the largest US power grid (Regional Transmission Organization), which serves 67 million customers in 13 states and Washington, DC, said its auction to procure power for the year starting June 2028 fell 6.8 gigawatts short of what it will need to guarantee system reliability during demand spikes, in a statement released Tuesday. The shortfall is equivalent to almost seven traditional nuclear reactors.

The result ramps up pressure on a grid that’s home to Virginia’s Data Center Alley, the biggest concentration of data centers in the US, and has borne the brunt of criticism for the struggle to manage the AI boom and sufficiently protect customers from soaring costs. Attention now shifts to an emergency procurement mechanism later this year that aims to shift the burden of ramping up power generation to hyperscalers.

6.831 Megawatt Shortfall

PJM Interconnection today announced the results of its 2028/2029 Base Residual Auction (BRA), which secured 138,318 MW of unforced capacity generation (UCAP) and demand response to meet projected electricity needs for the more than 67 million people across 13 states and the District of Columbia, which fall under the RTO’s umbrella.

Regions under the Fixed Resource Requirement (FRR) acquired an additional 10,864 MW in UCAP, for a total of 149,182 MW in UCAP available to serve forecasted peak electricity demand, plus a reserve margin. UCAP represents a generation resource’s maximum output adjusted for its estimated ability to reliably perform at times of highest system risk. The capacity of the resources procured in the auction, plus FRR resources, is short of PJM’s reliability requirement by 6,831 MW, meaning that the committed supply is less than what would be required to meet the one-event-in-10-year reliability standard (and with electricity-guzzling data centers popping up almost daily these days, the one-event-in-10-year has become a daily occurrence).

Keep reading

New York imposes the first statewide ban on large data centers

New York blocked the construction of any new large data centers for up to a year on Tuesday while the state creates rules to protect the environment and the energy grid from power-hungry facilities that fuel artificial intelligence.

Gov. Kathy Hochul signed an executive order imposing the country’s first statewide moratorium on hyperscale data centers, which house thousands of computer servers and require massive amounts of energy and a steady supply of water to keep cool.

The move puts the state in the center of a national debate over how to regulate the AI industry, as concerns over rising electric bills and environmental risks collide with the need to stimulate local economies and foster the U.S. tech sector.

“It’s my responsibility to take action and lead,” Hochul, a Democrat, said in a statement.

In effect, the executive order pauses state permitting for new large data centers and direct state regulators to create standards that address environmental impacts, energy demand, water usage and other factors, the governor’s office said.

President Donald Trump has warned states not to slap regulations on the AI industry, echoing tech companies in arguing such moves hamper job growth and cede ground to China in a race to lead in the rapidly growing field.

Earlier this year, Maine seemed poised to establish a similar moratorium. But the measure was vetoed by Democratic Gov. Janet Mills because it would have blocked a proposed data center in a town that has struggled after a mill closed.

Moratoriums have been proposed in at least a dozen states but have not gotten far, though some counties and municipalities have imposed their own temporary bans.

The decision in New York also carries political significance for Hochul’s reelection campaign and the state’s tight congressional races this fall, as Democrats move to address affordability concerns over high utility bills. In addition, the governor this year softened New York’s ambitious goals to reduce greenhouse gases, citing rising energy costs for consumers.

Hochul’s Republican opponent in the governor’s race, Nassau County Executive Bruce Blakeman, opposes a statewide moratorium and says local governments should be allowed to strike deals with tech companies for data center projects that promise enough economic benefits.

The state Legislature this year approved its own moratorium bill, but Hochul’s office described the legislation as complex and said it needed additional work. Instead, the governor opted for an executive order that would take effect immediately.

New York, at this stage, has not been a destination for the biggest hyperscale data centers.

Keep reading