Technology

By 2030, AI will greatly outperform humans in some complex intellectual tasks. Discover how LLMs are doubling their capabilities every seven months.

Use !friendica@lemmy.ca instead!

Some key insights from the article:

Basically, what they did was to look at how much batteries would be needed in a given area to provide constant power supply at least 97% of the time, and the calculate the costs of that solar+battery setup compared to coal and nuclear.

Microsoft has confirmed it is to axe another 9000 with its continued push into AI. I just find it incredible because for me AI is NOT the answer in any way shape or form.....

https://www.bbc.co.uk/news/articles/cdxl0w1w394o

Google AdSense announced a new feature to "Fill empty in-page ads" by showing "contextually-relevant suggestions related to your page and site inside unfilled ad spaces." In short, if an ad is not filled by Google on your page, then Google will look for other ways to monetize that ad spot.

I should note at the top, that Google will turn this on by default for all AdSense publishers within 30 days. So if you want, you can opt out of it.

This is pretty disturbing, I hope she gets a massive paycheck for this violation.

FOIA.

The Justice Department is advancing a radical theory of presidential power, nullifying Congress’s foreign affairs powers whenever the president finds them inconvenient.

These letters provide a legal rationale (if it can be called that) for the Trump administration’s commitment not to enforce the Protecting Americans from Foreign Adversary Controlled Applications Act (PAFACAA), the divestment-or-ban law that the Supreme Court upheld in January. The letters make two central claims, both of which are astonishing in their breadth and implications for executive power.

• Akamai Technologies
• Amazon
• Apple Inc.
• Digital Realty Trust, Inc.
• Fastly, Inc.
• Google, Inc.
• LG Electronics USA, Inc.
• Microsoft Corporation
• Oracle Corporation
• T-Mobile US, Inc.

Around half also believe Big Tech companies are more powerful than the EU

The Basque Country is implementing Quantus Skin in its health clinics after an investment of 1.6 million euros. Specialists criticise the artificial intelligence developed by the Asisa subsidiary due to its "poor” and “dangerous" results. The algorithm has been trained only with data from white patients.

• At the charging stations, daily concentrations of dangerous air particles, known as PM2.5, ranged from 7.3 to 39.0 micrograms per cubic meter.
• Urban sites without fast-charging stations had concentrations of PM2.5 ranging from only 3.6 to 12.4 micrograms per cubic meter.
• The tiny particles likely come from particle resuspension around Direct Current Fast Charging power cabinets. Cooling fans designed to prevent the electronics from overheating can also stir up dust and particles from internal surfaces.

Artificial intelligence (AI) is reshaping the criminal justice system. Law enforcement agencies are using it to predict crime, expedite response, and streamline routine tasks. One of the most promising applications can be found in body camera programs, where AI is transforming unmanageable archives of footage into active sources of insight.

AI can now analyze hundreds of hours of video in seconds. Early pilot programs suggest that these video-reviewing tools, when guided by human oversight, can uncover critical evidence that might otherwise be overlooked, reduce pretrial bottlenecks, and identify potential instances of officer misconduct. But these benefits come with risks. Absent clear guardrails, the same technologies could drift toward government overreach, blurring the line between public safety and state surveillance.

The line between public security and state surveillance lies not in technology, but in the policies that govern it. To responsibly harness AI and mitigate these risks, we recommend that agencies and policymakers:

• Establish and enforce clear use policies. Statewide rules for body camera use and AI governance ensure consistency across jurisdictions, particularly in areas like body camera activation, evidence sharing, and public disclosure.
• Pair technology with human oversight. AI should enhance—not replace—human decision-making. Final judgments must rest with trained personnel, supported by independent policy oversight from civilian review boards.
• Safeguard civil liberties. Safeguards must be in place to protect individual rights, limit surveillance overreach, and ensure data transparency. For example, limiting facial recognition during constitutionally protected activities like protests will help ensure AI is aligned with democratic ideals.

With the right guardrails in place, AI can elevate body cameras from after-action archival tools to always-on intelligence tools, informing decisions in the moment, when it matters most.

• About This Project.
• Executive Summary.