cross-posted from: https://lemmygrad.ml/post/10028972

cross-posted from: https://news.abolish.capital/post/12938

American tech giant Meta Platforms is reportedly using an open-source artificial intelligence model developed by Alibaba Group Holding to reinvigorate its faltering AI efforts in another likely win for Chinese AI. According to a Bloomberg report on Wednesday, Facebook owner Meta was using Alibaba’s Qwen model, along with other open-source models from Google and OpenAI, as part of the training process for a new model code-named Avocado, which was expected to be released in the spring. The report...

From China - South China Morning Post via This RSS Feed.

cross-posted from: https://news.abolish.capital/post/12809

At a time when the American public, and especially Democratic voters, express overwhelming distrust of artificial intelligence and Big Tech, the top House Democrat is being accused of failing to meet the moment.

On Tuesday, in preparation for an executive order to be signed this week by President Donald Trump, which would seek to block states from implementing new AI regulations, House Minority Leader Hakeem Jeffries (D-NY) unveiled his own effort to cozy up to the industry, whose major players have set aside more than $200 million to push out anti-AI politicians during the 2026 midterms, according to the New York Times.

Jeffries announced the creation of a “House Democratic Commission on AI and the Innovation Economy,” which will “develop policy expertise in partnership with the innovation community, relevant stakeholders, and committees of jurisdiction.”

What immediately caught the eye of critics was the list of fellow Democrats Jeffries picked to serve on the commission. It will be co-chaired by Reps. Ted Lieu (Calif.), Josh Gottheimer (NJ), and Valerie Foushee (NC), with Reps. Zoe Lofgren (Calif.) and Frank Pallone (NJ) serving as ex officio co-chairs.

As Sludge reported Tuesday: "The panel’s leaders rank among the House Democrats with the deepest ties to Big Tech and AI, from holding millions of dollars in tech stock to the contributions they’ve raised for their campaigns and the Republican-backed deregulation bills they've signed onto."

In July, Gottheimer introduced a bill along with Rep. French Hill (R-Ark.) "that would require financial regulators to create 'AI Innovation Labs' where firms could experiment with AI-driven financial products under looser regulations and without the normal threats of enforcement actions."

Gottheimer is also a major stakeholder in Microsoft, which has invested tens of millions of dollars into AI and nearly $7.5 million on lobbying in 2025 so far. Beyond the almost $100,000 in contributions Gottheimer has received from Microsoft, he is also a former executive who received anywhere from $1 million to $5 million last year from his stock holdings in the company, according to financial disclosure forms. He also frequently trades in other AI power players like Amazon, Meta, and Dell.

Lofgren, meanwhile, has accepted more money from the Internet industry over the course of her career than all but one other current House Democrat—including $265,000 from Google, $115,000 from Apple, and $110,000 from Meta, according to data from OpenSecrets.

In September 2024, Lofgren co-sponsored a bill introduced by Rep. Jay Abernolte (R-Calif.) which "would create a federal 'center for AI advancement and reliability' that it would instruct to work closely with private companies and other stakeholders on developing 'voluntary best practices and technical standards for evaluating the reliability, robustness, resilience, security, and safety of artificial intelligence systems.'"

Foushee, a member of the corporate-backed New Democrat Coalition, rode to Congress in 2022 with more than $1 million from the Protect Our Future political action committee, which was backed by former FTX CEO and convicted fraudster Sam Bankman-Fried.

In response to Trump's industry-friendly "AI Action Plan" in July, Foushee and the New Democrats unveiled their own "Innovation Agenda," which called for federal tax credits to companies that "reskill" workers and perform private research and development as well as federal investments in apprenticeships and "labor market data modernization."

Jeffries has neglected to take a position on Trump's proposal to preempt state regulations. Last Monday, he told reporters, "That conversation hasn't been brought to the leadership level yet."

In his statement announcing the Democratic commission on Tuesday, Jeffries said, "It is important that American companies continue to thrive" in the arena of AI, while "at the same time, Congress must consider what policies are needed to prevent bad actors from exploiting this transformative technology and inflicting harm upon the American people." However, he did not specifically mention Trump's pending block on state regulations.

— (@)

A poll released Friday by the progressive group Demand Progress showed that Americans across the political spectrum are unsettled by AI's influence in Washington: 68% of respondents overall said they were more worried that "the US government will not regulate artificial intelligence enough," as opposed to just 21% who feared too much regulation. While Democrats and independents were somewhat more concerned about underregulation at 71%, Republicans largely shared those fears, with 62% saying they feared the government would not regulate AI enough.

The consensus was even stronger regarding Big Tech's power over AI policy, with 78% of respondents overall saying it had too much influence. This included 81% of Democrats and independents and 74% of Republicans.

With this in mind, many critics were puzzled by Jeffries' decision to stack his AI commission with some of the industry's top allies.

— (@)

As Aaron Regunberg wrote in the New Republic last month, harnessing anger against the rapid, largely unregulated expansion of expensive, energy-sucking AI data centers was an essential part of Democrats' victories across the board in November's off-year elections:

In New Jersey, Gov.-elect Mikie Sherrill’s closing argument was a pledge to freeze electricity rates, which have soared because of data-center demand.

In Virginia, Gov.-elect Abigail Spanberger won after pledging to make data centers “pay their own way,” and many Democrats went even further.

At least one candidate, John McAuliff, flipped a seat in the House of Delegates by focusing almost entirely on tying his Republican opponent to the “unchecked growth” of data centers, with an ad that asked, “Do you want more of these in your backyard?”

And in Georgia, Democrats won their first nonfederal statewide races in decades, earning 60% of the vote against two Republican members of the Public Service Commission by criticizing Big Tech “sweetheart deals” and campaigning for policies “to ensure that the communities that they’re extracting from” don’t end up with their “water supplies … tapped out or their energy … maxed out.”

"This is the most populist moment of voter rage I've ever seen, and the leading Democrats are absolutely hostile to the idea of doing anything to address Silicon Valley's massive power," said Matt Stoller, an anti-monopoly expert.

"Anticorruption is one of the strongest arguments with the broadest appeal in American politics right now, but the Democratic leadership simply refuses to stop tanking it," added Matt Duss, a former advisor to Sen. Bernie Sanders (I-Vt.).

"I have never seen a gulf this wide between Democratic leadership and the party writ large," said author Zachary D. Carter. "The top is corrupt, the base is raging against corruption."

From Common Dreams via This RSS Feed.

Battery prices have plunged to a shocking new low of $99 per kWh, marking the cheapest level ever recorded. The drop is driven largely by the rapid dominance of LFP chemistry, boosted by massive scale, cheaper materials, and China-led manufacturing efficiency. This milestone could accelerate global EV adoption and slash production costs across the industry.

The paper argues that we are hitting a wall with current AI because we are obsessed with number crunching instead of structure.

Belabes posits that modern AI is too focused on statistical minimization and processing speed, which reduces everything to collections of numbers that inherently lack meaning. You lose the essence of what you are actually trying to model when you strip away the context to get raw data. The author suggests a pivot to Alexandre Grothendieck's Topos theory, which provides a mathematical framework for understanding geometric forms and preserving the deep structure of data rather than just its statistical number crunching.

Topos theory focuses on finding a new style space that acts as a bridge between different mathematical objects. Instead of just looking at points in a standard space, a topos allows us to look at the relationships and sheaves of information over that space, effectively letting us transfer invariants from one idea to another. It creates a way to connect things that seem totally unrelated on the surface by identifying their common essence. Belabes links this to the idea of conceptual strata where something that looks like noise or insignificant data in one layer might actually be critical structure in another layer. It's a move away from the binary notion of significant versus insignificant data and toward a relativistic view where significance depends on the conceptual layer you are analyzing.

The author uses literary examples like Homer and Dostoevsky to show that authentic meaning often precedes the words used to express it, whereas our current digital systems treat language as a closed loop where words define other words. Current AI essentially simulates discourse without the underlying voice or intent. By adopting a Topos-based approach, we might be able to build systems that respect these layers of meaning and read slowly to extract the actual shape of the information. It is basically a call to stop trying to brute force intelligence with bigger matrices and start modeling the actual geometry of thought.

cross-posted from: https://lemmy.zip/post/54662795

'User Alignment Critic' will review agentic actions so bots don't do things like emptying your bank account