AI Watch

Blurb:

In the early 2010s, nearly every STEM-savvy college-bound kid heard the same advice: Learn to code. Python was the new Latin. Computer science was the ticket to a stable, well-paid, future-proof life.

But in 2025, the glow has dimmed. “Learn to code” now sounds a little like “learn shorthand.” Teenagers still want jobs in tech, but they no longer see a single path to get there. AI seems poised to snatch up coding jobs, and there aren’t a plethora of AP classes in vibe coding. Their teachers are scrambling to keep up.

“There’s a move from taking as much computer science as you can to now trying to get in as many statistics courses” as possible, says Benjamin Rubenstein, an assistant principal at New York’s Manhattan Village Academy. Rubenstein has spent 20 years in New York City classrooms, long enough to watch the “STEM pipeline” morph into a network of branching paths instead of one straight line. For his students, studying stats feels more practical.

Blurb:

From The Guardian: “The use of artificial intelligence in healthcare could create a legally complex blame game when it comes to establishing liability for medical failings, experts have warned.

The development of AI for clinical use has boomed, with researchers creating a host of tools, from algorithms to help interpret scans to systems that can aid with diagnoses. AI is also being developed to help manage hospitals, from optimising bed capacity to tackling supply chains.

But while experts say the technology could bring myriad benefits for healthcare, they say there is also cause for concern, from a lack of testing of the effectiveness of AI tools to questions over who is responsible should a patient have a negative outcome.

AMD: Latest news and insights | Network World

AMD: Latest news and insights | Network World

AMD wins massive AI chip deal from OpenAI with stock sweetener– arstechnica.com
Source Link
Excerpt:

As part of the arrangement, AMD will allow OpenAI to purchase up to 160 million AMD shares at 1 cent each throughout the chips deal.

With demand for AI compute growing rapidly, companies like OpenAI have been looking for secondary supply lines and sources of additional computing capacity, and the AMD partnership is part the company’s wider effort to secure sufficient computing power for its AI operations. In September, Nvidia announced an investment of up to $100 billion in OpenAI that included supplying at least 10 gigawatts of Nvidia systems. OpenAI plans to deploy a gigawatt of Nvidia’s next-generation Vera Rubin chips in late 2026.

OpenAI has worked with AMD for years, according to Reuters, providing input on the design of older generations of AI chips such as the MI300X. The new agreement calls for deploying the equivalent of 6 gigawatts of computing power using AMD chips over multiple years.

Beyond working with chip suppliers, OpenAI is widely reported to be developing its own silicon for AI applications and has partnered with Broadcom, as we reported in February. A person familiar with the matter told Reuters the AMD deal does not change OpenAI’s ongoing compute plans, including its chip development effort or its partnership with Microsoft.

Lockheed, Verizon testing 5G-linked drone swarm for intel collection

Lockheed, Verizon testing 5G-linked drone swarm for intel collection

‘Swarms of Killer Robots’: Why AI is Terrifying the American Military – Politico
Source Link
Excerpt:

Artificial intelligence technology is poised to transform national security. In the United States, experts and policymakers are already experimenting with large language models that can aid in strategic decision-making in conflicts and autonomous weapons systems (or, as they are more commonly called, “killer robots”) that can make real-time decisions about what to target and whether to use lethal force.

But these new technologies also pose enormous risks. The Pentagon is filled with some of the country’s most sensitive information. Putting that information in the hands of AI tools makes it more vulnerable, both to foreign hackers and to malicious inside actors who want to leak information, as AI can comb through and summarize massive amounts of information better than any human. A misaligned AI agent can also quickly lead to decision-making that unnecessarily escalates conflict.

Risks and Benefits of AI for Businesses and Cybersecurity | SBS

Risks and Benefits of AI for Businesses and Cybersecurity | SBS

As AI redefines work, US employers cut jobs and remain cautious in hiring – Computerworld– www.computerworld.com
Source Link
Excerpt:

 

Ben Johnston, COO of small business lender Kapitus, said US businesses are grappling with tariffs that raise costs on imported goods, potentially making domestic manufacturing more competitive long term. But in the short term, those tariffs risk driving up inflation and disrupting global supply chains, threatening jobs across the manufacturing, wholesale, and retail sectors.

AI is also beginning to displace workers, especially in white-collar jobs. Companies are currently investing heavily in AI technologies that can analyze data and quickly make decisions that once could only be made by humans, Johnston said.

Companies are using AI to gather and analyze data from the web, internal systems, and third parties — tasks once done only by humans — mainly in white-collar roles like analytics and underwriting. And as robotics advance, AI could soon take on physical tasks in blue-collar jobs like driving, factory work, and even home healthcare, Johnston noted.

How U.S.-Gulf AI Deals Project Power– warontherocks.com
Source Link
Excerpt:

The great-power contest is not unfolding on battlefields or carrier decks, but inside data halls cooled by air conditioning, far from America’s shores. Rows of servers and racks of graphics processing units now carry as much strategic weight as military bases once did. Each deal for cloud access or advanced chips is a form of statecraft, binding partners into one camp’s technology ecosystem while locking out the other.

The United States is using AI infrastructure — data centers, cloud controls, and compute access — as a tool of power projection in the Arabian Gulf. By tying investment and capacity to governance safeguards, Washington can align regional partners with its security preferences, crowd out Chinese platforms, and set the rules for how AI is built and deployed. But the leverage is fragile. Without resilience and enforceable compliance, these arrangements risk becoming single points of failure or, worse, conduits for adversaries.

To make this new form of statecraft durable, U.S. policymakers should establish standard deal architectures with Gulf partners that combine hard technical safeguards, strict governance requirements, and built-in contingency plans. That means binding model weights to secure enclaves, tracking accelerators and workloads, embedding snapback clauses for violations, and pairing technical assurances with human rights standards. Done right, this approach can turn American-backed AI infrastructure into a lasting source of influence — quiet, scalable, and harder to dislodge than a forward operating base.

  1. A Feminist Approach to AI in Sub-Saharan Africa • Stimson Center– www.stimson.org
    Source Link
    Excerpt:

    Across Africa, AI is being harnessed to achieve positive impacts for marginalized communities. However, while AI can be used for good, some fear it could further marginalize and harm those it is intended to empower. Despite emphasis from both public and private sectors on equality and equity, uncertainty around policy-enabling environments, skills, and resources still presents a bottleneck for building inclusive AI. Though there are promising femtech solutions aimed at addressing specific gender concerns, the question of addressing needs and wants from a feminist approach in AI lingers.

    Gender is often still an afterthought when it comes to policy implementation and practice, but there are many initiatives, charters, and agreements in Africa that support equality and aim to eliminate violence against women, combat the disproportionate effect of poverty on women, and support women’s participation in the political and economic spheres. For example, Agenda 2063 promotes gender equality and an engaged, empowered youth. The African Union strategy on Gender Equality and Women’s Empowerment (GEWE) 2018-2028 also aims to strengthen women’s agency in Africa and ensure that women’s voices are amplified and their concerns are fully addressed. The African Charter on Human and Peoples’ Rights on the Rights of Women in Africa similarly requires member states to tackle “all forms of discrimination against women through appropriate legislative measures.”

    In Africa, there is a strong normative framework on gender equality and women’s and girls’ rights, and correspondingly, there are some civil or governmental initiatives that support women in national digital transformation policies. In Rwanda, for example, women have become increasingly influential in building national plans for artificial intelligence, and in Kenya, women have impacted the national plan for data uses. However, Africa is still facing challenges in integrating AI and policy. The Oxford AI Worldwide Readiness Index exemplifies the gap between the United States, ranked as first, and Mauritius, considered the African flagship country in AI policy, ranked 69th. There are only four African countries – Mauritius, South Africa, Rwanda, and Egypt – whose scores were higher than the global average of 47.59.

    Policy efforts across the continent are increasing but still limited. In mid-2021, Egypt launched its national AI strategy, christened “Artificial Intelligence for Development and Prosperity,” making clear the country’s ambitious goals for development and economic growth. Senegal followed suit in 2023 with a strategy of its own, also focused on economic development. On April 20, 2023, Rwanda released its “National AI Policy for Responsible AI Adoption,” which emphasizes AI for sustainable development. In 2024, Kenya published its draft national AI strategy with goals including social inclusion, ethics, and equity in AI.

     

The Complex Promise and Perils of AI in Policing : Center for ...

The Complex Promise and Perils of AI in Policing : Center for ...

Arizona’s AI policing tool threatens civil liberties– www.theblaze.com
Source Link
Excerpt:

Several Arizona police departments are piloting a new AI-powered policing tool that promises to revolutionize how officers catch criminals. But without robust constitutional safeguards, this cutting-edge technology could pose a serious threat to the civil liberties of everyday Americans.

Arizona police agencies are now testing a new AI program that “deploys lifelike virtual agents, which infiltrate and engage criminal networks across various channels.” The program, called Overwatch, was developed by Massive Blue and provides police departments with up to 50 different AI personas.

While the technology could, in theory, be used for noble purposes, … it also creates new opportunities for government overreach.

These include a sex trafficker persona, an escort persona, a 14-year-old boy in a child trafficking scenario, and a vaguely defined “college protester.” Beyond social media monitoring, the program allows police to communicate directly with suspects while posing as one of these AI-generated personas, all without a warrant.

No transparency

So far, both the police departments using Overwatch and the company behind it have been extremely secretive about its operations. Massive Blue co-founder Mike McGraw declined to answer questions from 404 Media, which first broke the story, about how the program works, which departments are using it, and whether it has led to any arrests.

“We cannot risk jeopardizing these investigations and putting victims’ lives in further danger by disclosing proprietary information,” McGraw said.

The Pinal County Sheriff’s Office, one of the few agencies that have confirmed using the program, admitted it has not yet led to any arrests. Officials refused to provide details, saying, “We cannot risk compromising our investigative efforts by providing specifics about any personas.”

At an appropriations hearing, a Pinal County deputy sheriff also declined to share information about the program with the county council. Remarkably, the Arizona Department of Public Safety, which funds the initiative, does not appear to have been informed about the program’s specifics.

While the technology could, in theory, be used for noble purposes, such as preventing terrorist attacks or combating human trafficking, it also creates new opportunities for government overreach. Without safeguards, it poses a direct threat to the civil liberties of innocent Americans.

Invitation to entrapment

History is full of examples of government entrapment and abuse of power. In the plot to kidnap Michigan Gov. Gretchen Whitmer (D-Mich.), for example, FBI involvement played a central role in bringing groups together that may never have otherwise connected.

Similarly, in Jacobson v. United States (1992), federal agents sent child sexual abuse material through the mail to a man with no prior criminal record, leading to his conviction, which was later overturned.

RELATED: Netflix’s chilling new surveillance tools are watching you

Photo Illustration by Piotr Swat/SOPA Images/LightRocket via Getty Images

In both cases, it is doubtful the crimes would have occurred without government intervention. A program like Overwatch makes such abuses easier, granting the government new ways to monitor and manipulate citizens who have never been convicted of a crime, and all without warrants.

The risks are compounded by the program’s vague and troubling categories, such as “college protester,” which could be redefined depending on who is in power. That opens the door for the technology to be weaponized against political dissent, even when no crime has been committed.

Without serious constitutional safeguards, programs like this are poised to become political tools of tyranny. Americans must demand warrant requirements and legislative oversight before this technology spreads nationwide and the erosion of our constitutional liberties becomes irreversible.

California Governor Gavin Newsom signs landmark AI safety law SB 53– fortune.com
Source Link
Excerpt:

California has taken a significant step toward regulating artificial intelligence with Governor Gavin Newsom signing a new state law that will require major AI companies, many of which are headquartered in the state, to publicly disclose how they plan to mitigate the potentially catastrophic risks posed by advanced AI models.

The law also creates mechanisms for reporting critical safety incidents, extends whistleblower protections to AI company employees, and initiates the development of CalCompute, a government consortium tasked with creating a public computing cluster for safe, ethical, and sustainable AI research and innovation. By compelling companies, including OpenAI, Meta, Google DeepMind, and Anthropic, to follow these new rules at home, California may effectively set the standard for AI oversight.

Newsom framed the law as a balance between safeguarding the public and encouraging innovation. In a statement, he wrote: “California has proven that we can establish regulations to protect our communities while also ensuring that the growing AI industry continues to thrive. This legislation strikes that balance.”

The legislation, authored by State Sen. Scott Wiener, follows a failed attempt to pass a similar AI law last year. Wiener said that the new law, which was known by the shorthand SB 53 (for Senate Bill 53), focuses on transparency rather than liability, a departure from his prior SB 1047 bill, which Newsom vetoed last year.

“SB 53’s passage marks a notable win for California and the AI industry as a whole,” said Sunny Gandhi, VP of Political Affairs at Encode AI, a co-sponsor SB 53. “By establishing transparency and accountability measures for large-scale developers, SB 53 ensures that startups and innovators aren’t saddled with disproportionate burdens, while the most powerful models face appropriate oversight. This balanced approach sets the stage for a competitive, safe, and globally respected AI ecosystem.”

‘Say it to my face’: House minority leader slams Trump; calls US president ‘racist’ over deepfake video– timesofindia.indiatimes.com
Source Link
Excerpt:

 

House Minority Leader Hakeem Jeffries on Tuesday accused President Donald Trump of being “racist” over a deepfake video posted to the social media platform, which depicted him wearing a sombrero.The fabricated video was posted on Trump’s Truth Social account on Monday night and featured Jeffries sporting an exaggerated handlebar moustache with mariachi music playing in the background. The video, released hours after Trump met with congressional leaders about the looming government shutdown, appeared aimed at mocking Democrats’ stance on federal benefits for migrants — a claim Democrats deny. Speaking on the steps of the Capitol, where numerous Democrats rallied for Republicans to negotiate a bipartisan spending bill, Jeffries stated, “Mr. President, the next time you have something to say about me, don’t cop out through a racist and fake AI video.” He added, “When I’m back in the Oval Office, say it to my face. Say it to my face.”Trump, known for attacking minorities and exacerbating racial tensions to energise his MAGA base, has a longstanding record that Jeffries, the first Black lawmaker to lead either party, has consistently opposed. Uploaded hours after Trump met with congressional leaders regarding the shutdown, the video parodied a press conference held by Jeffries and Senate Minority Leader Chuck Schumer of New York outside the White House following that meeting. The fake audio portrayed Schumer attacking his own party as being deeply unpopular “because of all of our woke, trans bulls—.” It further asserted, “Not even Black people want to vote for us anymore.

Major advance: Stanford researchers use AI to design viruses; warn of dangers if technology misused– timesofindia.indiatimes.com
Source Link
Excerpt:

A research team in California has used artificial intelligence to design viral genomes, which were later built and tested in a laboratory. Some of these AI-created viruses successfully infected bacteria, demonstrating that generative AI models can produce functional genetic material.Researchers at Stanford University and the Arc Institute in Palo Alto described the achievement as “the first generative design of complete genomes.” Jef Boeke, a biologist at NYU Langone Health, called it a major step toward AI-designed life forms, according to Newsweek citing MIT Technology Review.Boeke noted that the viruses displayed new genes, shortened genes and even different gene arrangements compared to natural versions.The team used their AI system, called Evo, a large language model similar to ChatGPT to design 302 complete genomes.

AI watchdog’s chilling ‘everyone will die’ prediction with robot update in less than five years– www.mirror.co.uk
Source Link
Excerpt:

Members of an AI watchdog group have warned in a new book that computational systems could eventually become self-aware to an extent that would be deadly to humanity

Artificial Intelligence could pose a real threat to humanity, a group has warned(Image: Getty Images)

Superpowered Artificial Intelligence could bring about the extinction of humankind in just a few years, researchers have claimed.

AI risk researchers have joined together to issue an urgent warning about the future of AI in a new book, titled If Anyone Builds It, Everyone Dies, claiming a frightening iteration of the already advanced technology could arrive imminently. They claim Artificial superintelligence (ASI) is just two to five years from development, and that it spells doom for humanity.

When it arrives, they have sensationally claimed that “everyone, everywhere on Earth, will die”, and warned people spooked by the research they should join the call to have development paused “as soon as we can for as long as necessary”.

California Lawyer Punished for Using AI That Cited Fake Cases– www.breitbart.com
Source Link
Excerpt:

A lawyer in California was hit with sanctions by the courts after it was discovered that he had been citing cases that did not exist — and that were completely made up by artificial intelligence (AI) programs.

The Daily Journal reported:

By now, most attorneys are familiar with the limitations of artificial intelligence–particularly its tendency to produce fabricated citations and unreliable authority–but the California Court of Appeal has now made clear that those “hallucinations” can carry very real consequences.

In Noland v. Land of the Free, L.P. (B331918, Sept. 12, 2025), Division Three of the Second District sanctioned the plaintiff/appellant’s attorney $10,000 and reported him to the state bar after finding that nearly all of the legal quotations in his appellate briefs were invented by generative AI tools and had never appeared in any published case.

And the Court was clear as to the reasons why it was publishing the opinion: “We … publish this opinion as a warning. Simply stated, no brief, pleading, motion, or any other paper filed in any court should contain any citations–whether provided by generative AI or any other source–that the attorney responsible for submitting the pleading has not personally read and verified.” (emphasis in original).

Meta AI Dings Kirk for Supposed Misinformation, Citing ‘Neutral’ Wikipedia– www.newsbusters.org
Source Link
Excerpt:

Meta AI went after Turning Point USA Founder Charlie Kirk by relying on none other than Wikipedia, exposing the disturbing influence of the online encyclopedia on artificial intelligence.

On Wednesday, the chatbot was quick to accuse Kirk of promoting conspiracy theories and other purportedly false claims when asked, “Who is Charlie Kirk?” Worse still, Meta AI identified Wikipedia as its sole source, justifying its choice by stating it offers a “neutral point of view and comprehensive coverage of current events.”

 

The Download: AI-designed viruses, and bad news for the hydrogen industry– www.technologyreview.com
Source Link
Excerpt:

Artificial intelligence can draw cat pictures and write emails. Now the same technology can compose a working genome.

A research team in California says it used AI to propose new genetic codes for viruses—and managed to get several of them to replicate and kill bacteria.

The work, described in a preprint paper, has the potential to create new treatments and accelerate research into artificially engineered cells. But experts believe it is also an “impressive first step” toward AI-designed life forms. Read the full story.

—Antonio Regalado

Clean hydrogen is facing a big reality check

Hydrogen is sometimes held up as a master key for the energy transition. It can be made using several low-emissions methods and could play a role in cleaning up industries ranging from agriculture to aviation to shipping.

This moment is a complicated one for the green fuel, though, as a new report from the International Energy Agency lays out. A number of major projects face cancellations and delays. The US in particular is seeing a slowdown after changes to key tax credits and cuts in support for renewable energy.

Still, there are bright spots for the industry, including in China, and new markets could soon become crucial for growth. Here are three things to know about the state of hydrogen in 2025.

—Casey Crownhart

 

International Students in China Complain, “Quark AI Has Forgotten Us!”– chinadigitaltimes.net
Source Link
Excerpt:

Some international students in China have taken to social media platform RedNote (Xiaohongshu) to complain about being excluded from obtaining free educational accounts for Quark AI, an LLM tool widely used by their Chinese university classmates. Using the hashtag #WeStudyInChina, these students have also set up an online “message wall” to lobby for inclusion in the popular AI tool.

Chinese online reactions to the students’ pleas ranged from sympathy to amusement, Schadenfreude to national pride. Some commenters highlighted the perceived privileges enjoyed by exchange students in China, while others pointed to the clamor for Quark AI as a sign that Chinese AI tools have finally become cutting edge. Others noted the similarities between these recent “Quark AI refugees” and the millions of so-called “TikTok refugees” who joined RedNote earlier this year when a U.S. ban on TikTok seemed imminent. (After U.S.-China bilateral trade talks in Spain this weekend, U.S. Treasury Secretary Scott Bessent said that the two nations had reached a framework deal to divest TikTok’s ownership from Chinese parent company Bytedance.)

In a recent WeChat article titled “As International Students in China Are Reduced to ‘Quark AI Refugees,’ Should We Gloat, or Feel Proud?” blogger Xiang Dongliang discusses various aspects of the controversy, including China’s educational subsidies for foreign students, on-campus segregation of Chinese and international students, and the rise of xenophobic attitudes in Chinese society. In the end, Xiang suggests that neither smug nationalism nor petty Schadenfreude is an appropriate reaction to the travails of exchange students in China:

I came across this particularly interesting trending topic: “Quark AI has forgotten us!”

Secrets of Chinese AI Model DeepSeek Revealed in Landmark Paper– www.scientificamerican.com
Source Link
Excerpt:

The success of DeepSeek’s powerful artificial intelligence (AI) model R1 — that made the US stock market plummet when it was released in January — did not hinge on being trained on the output of its rivals, researchers at the Chinese firm have said. The statement came in documents released alongside a peer-reviewed version of the R1 model, published today in Nature.

R1 is designed to excel at ‘reasoning’ tasks such as mathematics and coding, and is a cheaper rival to tools developed by US technology firms. As an ‘open weight’ model, it is available for anyone to download and is the most popular such model on the AI community platform Hugging Face to date, having been downloaded 10.9 million times.

The paper updates a preprint released in January, which describes how DeepSeek augmented a standard large language model (LLM) to tackle reasoning tasks. Its supplementary material reveals for the first time how much R1 cost to train: the equivalent of just US$294,000. This comes on top of the $6 million or so that the company, based in Hangzhou, spent to make the base LLM that R1 is built on, but the total amount is still substantially less than the tens of millions of dollars that rival models are thought to have cost. DeepSeek says R1 was trained mainly on Nvidia’s H800 chips, which in 2023 became forbidden from being sold to China under US export controls.

AI reveals how toughest protein bonds behave– cosmosmagazine.com
Source Link
Excerpt:

Proteins can form “catch-bonds” that tighten under force, much like a finger trap. Credit: Rafael C. Bernardi, Auburn Physics

Researchers have used artificial intelligence to help uncover how certain protein interactions act like a finger trap, gripping tighter the harder they are pulled.

These interactions, known as catch-bonds, are essential in how the body holds together under stress and how bacteria attach to cells.

The researchers suggest that a better understanding of these bonds could help inform the design of new medications and biomaterials.

Scientists have been unsure as to whether these catch-bonds activate straight away or if they need to be stretched to a certain threshold before they ‘switch on’.

The new study discovered that these bonds activate almost immediately after a force is applied.