Author: rb809rb

  • Google and the Pentagon sign classified deal to give the Department of Defense unfettered access to its AI models

    Google has signed a deal that allows the US Department of Defense to use its AI models for “any lawful government purpose.” This is according to a report by The Information, which also notes that the full details of the contract are classified.

    An anonymous source within the company has suggested that the two entities have agreed that the search giant’s AI tech shouldn’t be used for domestic mass surveillance or autonomous weapons “without appropriate human oversight and control.” However, the contract also reportedly doesn’t give Google “any right to control or veto” anything the government decides to do. In other words, the famously trustworthy US government will just have to be taken at its word.

    “We believe that providing API access to our commercial models, including on Google infrastructure, with industry-standard practices and terms, represents a responsible approach to supporting national security,” a Google spokesperson told Reuters. The spokesperson also echoed that the company holds the opinion that AI shouldn’t be used for mass surveillance or autonomous weaponry without appropriate human oversight. Some might argue that the technology shouldn’t be used for that stuff at all, oversight or not.

    To that end, nearly 600 Google employees just penned an open letter to CEO Sundar Pichai to urge the company against making this kind of deal with the Pentagon. This stems from concerns that the tech would be used in “inhumane or extremely harmful ways.”

    “Human lives are already being lost and civil liberties put at risk at home and abroad from misuses of the technology we are playing a key role in building,” the letter states. “As people working on AI, we know that these systems can centralize power and that they do make mistakes.”

    Google will join OpenAI and Elon Musk’s xAI in this endeavor, as they both have made classified AI deals with the US government. Anthropic had a deal in place, but refused the government’s demands to remove weapon and surveillance-related safeguards.

    That refusal annoyed President Trump and the Pentagon so much that Anthropic was entirely blacklisted from federal use. This doesn’t exactly sound like the actions of a government that is dedicated to “appropriate human oversight and control” of dangerous AI military tech. Engadget has reached out to Google to ask for more specifics and will update this post when we hear back.

  • A crypto coalition releases technical proposal to save Aave users from a massive token exploit

    A crypto coalition releases technical proposal to save Aave users from a massive token exploit

    A $300 million hole doesn’t usually come with a neat repair manual. This time, the group spearheading the Kelp DAO recovery effort is trying to write one.

    DeFi United, a coalition of multiple blockchain projects and crypto ecosystem individuals, has laid out a detailed, step-by-step plan to restore the backing of rsETH after this month’s Kelp DAO hack sent shockwaves through DeFi lending markets, releasing more than 116,000 tokens that weren’t properly accounted for.

    The proposal, circulated on Aave’s official X account, reads like a coordinated cleanup operation, one that leans heavily on Aave’s infrastructure to unwind the damage and get markets back on a stable footing.

    The incident traces back to April 18, when an attacker exploited a vulnerability in rsETH’s bridge. By forging a message that appeared legitimate, the attacker tricked the Ethereum side of the system into releasing 116,500 rsETH, making the system believe the funds had moved when they hadn’t, allowing a large batch of rsETH to be created without backing.

    Those tokens didn’t just sit idle. They were spread across multiple wallets and deployed across DeFi, with a significant portion used as collateral on Aave and other lending platforms.

    That’s where the problem became systemic: protocols like Aave suddenly found themselves holding collateral that, at least temporarily, wasn’t fully backed.

    According to the proposal, most of the exploited funds are still in play. Roughly 107,000 of the original 116,500 rsETH remain tied up in active positions across Aave and Compound.

    That leaves two problems to solve at once: restoring the actual backing of rsETH itself, and unwinding the loans created using those extra tokens.

    DeFi United’s proposal aims to tackle both sides of that equation simultaneously.

    On the backing side, the group says it has already lined up enough $ETH commitments to fully re-collateralize rsETH. The plan is to feed that $ETH back into the system in stages, converting it to rsETH and depositing it back into the system so the token is once again fully backed.

    At the same time, attention shifts to the lending markets where the damage is most visible.

    Instead of letting things play out chaotically, the plan is to step in and carefully unwind the mess.

    A big part of that involves dealing with the positions the attacker opened on Aave. These are essentially loans backed by rsETH that shouldn’t have existed in the first place. Rather than waiting for those loans to collapse on their own — which could cause more market disruption — the proposal suggests nudging the system so they can be closed out in a more controlled way.

    In practice, temporarily adjusting how rsETH is valued inside the system will enable those bad positions to be liquidated or closed more smoothly. As those positions are unwound, the underlying assets (like $ETH) can be recovered. The proposal estimates this could free up around 13,000 $ETH from Aave alone.

    Once that collateral is back in hand, it gets converted into $ETH and used to cover the shortfall created by the exploit — essentially filling the hole left behind.

    The process isn’t risk-free. It hinges on governance approvals across multiple chains, the successful deployment of committed funds and a smooth execution of the unwind.

    Still, the plan reflects a more coordinated response than DeFi has often managed previously. If executed as intended, the end goal is straightforward: “rsETH backing is fully restored, and all affected markets are stabilized,” as the proposal says.

    Read more: Industry leaders are pouring hundreds of millions into a rescue plan for Aave users after massive crypto hack

  • Google Translate uses AI to help you practice pronunciation

    Google is celebrating Translate’s 20th birthday by launching pronunciation practice, which the company says is one of the most requested features for the product. The feature is only rolling out on Android at the moment for English, Spanish and Hindi in the US and India. If it’s available for you, you’ll see a button at the bottom of the app that says “Practice,” which gives you the option to either “pronounce” what you’ve translated or to “listen” to how it’s actually pronounced by native speakers.

    If you choose the “pronounce” option, Translate will listen to you speak and then use artificial intelligence to analyze how you said the words to provide instance feedback. It will then show you a phonetic spelling of how specific words should be pronounced. In the example Google provided, for instance, the speaker pronounced the Spanish word for juice as “jugo” with the English “j” sound instead of with the Spanish “j” sound. So, Translate spells it out as “HU-go” in its pronunciation suggestion.

    Google said around third of users on mobile use Translate to practice speaking and listening in order to be able to hold real-world conversations, making this new feature a very useful addition. The company also revealed other stats about the app. Apparently, it now supports over 250 languages, including some endangered and indigenous ones, and has over 1 billion monthly user who have been translating over 1 trillion words every month.

  • Stuck kitten rescued from radiator in Wisconsin

    Stuck kitten rescued from radiator in Wisconsin

    Odd News // 3 weeks ago

    Wrong click earns Michigan woman a $251,738 lottery jackpot

    April 1 (UPI) — A Michigan woman said the wrong click at the right time while buying lottery tickets online led to her winning a $251,738 Fantasy 5 jackpot.

  • Ariana Grande Announces New Album ‘Petal’ and Sets Summer Release Date

    Ariana Grande Announces New Album ‘Petal’ and Sets Summer Release Date

    Ariana Grande is doing it all.

    After concluding her time as Glinda in the Wicked films with Wicked: For Good’s Thanksgiving release — and unveiling a couple more upcoming acting roles — she is also returning to music.

    On Tuesday, Grande took to Instagram to announce that her eighth studio album, Petal, is set to be released on July 31. The photo shows the Grammy winner without her signature ponytail, hinting at a new era coming.

    Petal, from Republic Records, is co-written by Grande and Swedish-Persian producer Ilya Salmanzadeh. On April 18, she shared a video on Instagram describing the album as “something that is full of life and growing through the cracks of something cold and hard and challenging.”

    Grande’s previous album, Eternal Sunshine, celebrated its three-year anniversary last month. The album was nominated for three Grammy Awards.

    On June 6, the singer will kick off her Eternal Sunshine tour in Oakland, Calf. The tour marks her first in seven years, following her last tour, which ended in December 2019, focused on her Sweetener and Thank U, Next albums.

    As far as acting, the Oscar-nominated actress’ upcoming projects include Focker In-Law, the fourth film of the Meet the Parents franchise, where she stars alongside Robert De Niro and Ben Stiller; Jon M. Chu’s Oh, the Places You’ll Go! animated feature adaptation opposite Josh Gad; and season 13 of Ryan Murphy’s American Horror Story, with a star-studded ensemble including Jessica Lange, Sarah Paulson, Evan Peters (who played Grande’s love interest in her “We Can’t Be Friends” music video), Angela Bassett, Emma Roberts (Grande’s Scream Queens co-star) and more.

  • NBA ID Postseason Challenge

    The 2026 NBA Playoffs are in full swing, filled with upsets, clutch moments and unheralded stars making an impact on the game’s brightest stage. NBA ID Members can join the action by taking part in the NBA ID Postseason Challenge, which runs from April 28 to May 3. It’s a limited-time opportunity to engage with the moments defining the postseason.


    How it works

    Complete the required activities in the NBA App or on NBA.com from April 28 to May 3 to earn exclusive postseason rewards, including badges, digital wallpapers and more. Members must be signed in to participate in the NBA ID Postseason Challenge. 

    Watch: Watch Stories or Moments, or tune into live or classic games for at least five minutes

    Play: Join in on any NBA Play game or put your fandom to the test in NBA Pick’Em Streak to the Finals 

    Vote: Cast a vote in any eligible NBA ID voting experience


    What you’ll earn

    • All Members: Complete the challenge to access an exclusive NBA ID Postseason Challenge Badge and score unique digital wallpapers packed with playoff visuals and Eastern vs. Western Conference hype. 
    • US & Canada Members: You’ll also be entered into a sweepstakes for a chance to win a trip for two to the 2026 NBA Finals.

    Terms & conditions apply* 


    Why Join?

    The NBA ID Postseason Challenge puts fans in the middle of the action, letting you vote on the biggest playoff moments and storylines shaping the postseason. Earn exclusive rewards just for participating as an NBA ID member.

    Don’t miss out—join the NBA ID Postseason Challenge – watch, play, vote, and earn rewards throughout the 2026 NBA Playoffs.

  • FIFA cut World Cup ref following arrest for alleged sexual assault in UK

    FIFA cut World Cup ref following arrest for alleged sexual assault in UK

    FIFA ‘aware of serious allegation’ following UK arrest of match official who will now not be listed for World Cup.

    A football match official who was a potential pick to work at the men’s World Cup in North America was arrested for an alleged sexual assault at a hotel while in Britain for a game.

    World Cup organiser FIFA said on Tuesday that it was “aware of the serious allegation”, after details of an alleged assault of a teenage boy were first reported by British daily The Sun.

    Recommended Stories

    list of 4 itemsend of list

    “In the meantime, the match official will not be considered for any FIFA competition matches,” the football governing body said in a statement.

    FIFA published a list this month of 52 referees, 88 assistant referees and 30 video match officials selected for the 104-game World Cup being co-hosted by the United States, Canada and Mexico from June 11 to July 19.

    The match official was not identified in reports, nor was the European competition game he was working at for UEFA.

    “We are monitoring the situation with great concern and will continue to follow developments closely,” UEFA said.

    UEFA added that it also will not be selecting the official for its games, and the alleged incident was part of “an active investigation”.

    The Sun reported that the Metropolitan Police in London released a man in his 30s on bail.

  • Elon Musk’s Fight With Colorado Over AI Law Hits Pause as State Considers Revisions

    Elon Musk’s Fight With Colorado Over AI Law Hits Pause as State Considers Revisions

    In brief

    • xAI and Colorado jointly moved to pause the lawsuit over SB24-205.
    • Enforcement of Colorado’s AI law is on hold while lawmakers consider amendments.
    • The case could resume if revisions fail to address xAI’s constitutional concerns.

    Colorado’s legal fight with Elon Musk’s artificial intelligence company xAI is on pause for now.

    In a joint filing on Friday, xAI and Colorado Attorney General Phil Weiser asked a federal court to cancel the June 16 scheduling conference and suspend all case deadlines in xAI’s lawsuit challenging Senate Bill 24-205, the state’s law aimed at preventing “algorithmic discrimination” in high-risk AI systems.

    The filing also temporarily halts enforcement of SB24-205, or any replacement law passed this legislative session. At the same time, Colorado lawmakers consider revisions, and the court weighs xAI’s expected motion for a preliminary injunction.

    Earlier this month, xAI sued Colorado seeking to block the state’s law before it takes effect. The company argues that SB24-205 would force developers to alter how AI systems operate and restrict how models generate responses.

    “SB24-205 is decidedly not an anti-discrimination law,” xAI’s attorneys wrote in the original complaint. “It is instead an effort to embed the State’s preferred views into the very fabric of AI systems.”

    The lawsuit argues the SB24-205 violates the First Amendment by forcing xAI’s chatbot, Grok, to answer certain questions in ways that match Colorado’s views on diversity and fairness. It also argues that the law is too unclear to enforce fairly, tries to regulate behavior outside Colorado, and treats some AI systems more favorably than others based on the kinds of answers they produce.

    The joint filing says a Colorado AI policy group formed by Gov. Jared Polis released a draft bill on March 17 to repeal and replace SB24-205. The attorney general said his office will not enforce the law or issue rules until the legislative session and rulemaking process are complete.

    Under the agreement, the attorney general said he will not launch enforcement actions or investigations against xAI for alleged violations until 14 days after the court rules on xAI’s expected injunction request.

    xAI agreed to file its motion for a preliminary injunction within 28 days after final adoption of rules implementing the law or any replacement measure.

    The legal fight escalated last week when the U.S. Department of Justice moved to intervene in support of xAI.

    The case is part of a broader fight over who should regulate artificial intelligence in the United States, as states including Colorado, New York, and California advance their own rules while the Trump administration pushes for a federal approach.

    Daily Debrief Newsletter

    Start every day with the top news stories right now, plus original features, a podcast, videos and more.

  • Trump’s FCC Preparing Review of ABC Broadcast Licenses Over Jimmy Kimmel’s Melania Joke: Report

    Trump’s FCC Preparing Review of ABC Broadcast Licenses Over Jimmy Kimmel’s Melania Joke: Report

    The FCC is gearing up to launch a formal review of the broadcast licenses granted to Disney’s eight owned ABC stations over complaints about Jimmy Kimmel‘s joke about Melania Trump looking like an “expectant widow,” coming days before a gunman crashed into the White House Correspondents’ Dinner with the apparent intent to try to assassinate President Trump, according to a report.

    The FCC, headed by Trump appointee Brendan Carr, “is moving toward a review of Disney’s broadcast licenses, according to people familiar with the matter, a maneuver that would up the pressure on the ABC owner as it faces fierce scrutiny from the administration — again — over a late night monologue,” per a report by Semafor.

    Carr has threatened to review — and potentially revoke — ABC’s broadcast TV licenses before, including earlier month when he criticized its DEI programs. “If the evidence does in fact play out and shows that they were engaged in race- and gender-based discrimination, that’s a very serious issue at the FCC, that could fundamentally go to their character qualifications to even hold a license,” Carr said in an interview with Fox News.

    Representatives for the FCC and ABC did not immediately respond to requests for comment.

    Kimmel’s joke on the Thursday, April 23, episode of “Jimmy Kimmel Live!” featured a mock White House Correspondents’ Dinner. During the bit, the host quipped that Melania had the glow of an “expectant widow” — prompting both President Trump and the First Lady to post separate statements on social media Monday calling for Disney and ABC to fire Kimmel.

    That came two days before the real WHCD on April 25 was thrown into chaos after an armed man charged through a security perimeter outside the event before he was apprehended. The suspect, identified by authorities as California resident Cole Tomas Allen, has been charged with attempting to assassinate Trump.

    On Monday’s episode of “Jimmy Kimmel Live!” the late-night host defended his remark about Melania as “a very light roast joke about the fact that he’s almost 80 and she’s younger than I am. It was not by any stretch of the definition a call to assassination. And they know that.”

    Kimmel continued, “I’ve been very vocal for many years speaking out against gun violence, in particular. But I understand that the First Lady had a stressful experience over the weekend. And probably every weekend is pretty stressful in that house. And also, I agree that hateful and violent rhetoric is something we should reject. I do. And I think a great place to start to dial that back would be to have a conversation with your husband about it.”

    Carr has threatened to use the FCC’s regulatory authority to revoke licenses of TV broadcasters that air programming that he finds objectionable — including over Kimmel’s on-air comments last September about Charlie Kirk’s assassination.

    Carr had threatened to investigate TV stations for “news distortion” if they didn’t drop Kimmel (following Kimmel’s comments about MAGA trying to score political points from Kirk’s killing). The FCC chairman has previously claimed he’s not censoring anybody and that he isn’t against free speech — he’s just against “fake news,” and he’s just doing his job to hold spectrum licensees accountable to a “public interest” standard.

    He’s dangled the same kind of threat in other instances, including last month when Carr implied the FCC would not renew licenses of broadcasters that perpetrated “hoaxes and news distortions” in their coverage of the Trump administration’s Iran war.

    That said, any charge the FCC lodged against a TV broadcast company about alleged “news distortion” would be tied up in bureaucratic proceedings for months or even years, even before it reached a court of law — where it would likely be challenged.

    On Monday, Trump blasted Kimmel over the Melania sketch and called for his firing.

    “Jimmy Kimmel, who is in no way funny as attested to by his terrible Television Ratings, made a statement on his Show that is really shocking,” Trump posted on Truth Social. “He showed a fake video of the First Lady, Melania, and our son, Barron, like they were actually sitting in his studio, listening to him speak, which they weren’t, and never would be. He then stated, ‘Our First Lady, Melania, is here. Look at Melania, so beautiful. Mrs. Trump, you have a glow like an expectant widow.’ A day later, a lunatic tried entering the ballroom of the White House Correspondents Dinner, loaded up with a shotgun, handgun, and many knives. He was there for a very obvious and sinister reason. I appreciate that so many people are incensed by Kimmel’s despicable call to violence… Jimmy Kimmel should be immediately fired by Disney and ABC.”

    Melania Trump, hours before the president weighed in, had written in a post on X, “Kimmel’s hateful and violent rhetoric is intended to divide our country. His monologue about my family isn’t comedy — his words are corrosive and deepens the political sickness within America,” Melania Trump posted on X hours before her husband’s own post.”

    She continued, “People like Kimmel shouldn’t have the opportunity to enter our homes each evening to spread hate. A coward, Kimmel hides behind ABC because he knows the network will keep running cover to protect him. Enough is enough. It is time for ABC to take a stand. How many times will ABC’s leadership enable Kimmel’s atrocious behavior at the expense of our community.”