news aggregator

You get a superintelligence and you get a superintelligence. Everybody gets a superintelligence

Meta is plowing tens of billions of dollars into GPU bit barns the size of Manhattan Island, and yet The Social Network has struggled to upstage rivals like OpenAI or Anthropic.…

India launched the $1.5 billion NISAR radar imaging satellite on Wednesday from the Satish Dhawan Space Centre, marking the first joint mission between NASA and the Indian Space Research Organisation. The satellite uses dual radar frequencies -- NASA's L-band and ISRO's S-band -- to detect Earth surface changes as small as one centimeter from its 747-kilometer orbit. NISAR will map the entire planet every 12 days using a 240-kilometer-wide radar swath, providing data for climate monitoring and disaster response that will be freely available to users worldwide.

Read more of this story at Slashdot.

IBM report shows a rush to embrace technology without safeguarding it, and as for governance...

Organizations rushing to implement AI are neglecting security and governance, IBM claims, with attackers already taking advantage of lax protocols to target models and applications.…

Dropbox will shut down its password manager service by October 28, giving users until then to extract their data before permanent deletion. The discontinuation occurs in phases: Dropbox Passwords becomes view-only on August 28, the mobile app stops working September 11, and complete shutdown follows October 28. The company cited focusing on core product features as the reason for dropping the service, which launched in 2020 for paid users and expanded to all users in 2021.

Read more of this story at Slashdot.

Data stream from aging sensor to continue after public backlash and amateur workaround

The US Navy has announced plans to continue distributing satellite data needed for hurricane forecasting, months after authorities said the data stream was to be turned off.…

Google will soon cast an even wider net with its AI age estimation technology. From a report: After announcing plans to find and restrict underage users on YouTube, the company now says it will start detecting whether Google users based in the US are under 18. Age estimation is rolling out over the next few weeks and will only impact a "small set" of users to start, though Google plans on expanding it more widely. The company says it will use the information a user has searched for or the types of YouTube videos they watch to determine their age. Google first announced this initiative in February. If Google believes that a user is under 18, it will apply the same restrictions it places on users who proactively identify as underage.

Read more of this story at Slashdot.

Read-only in weeks, deleted forever in months

Dropbox has given users of its password manager until the end of October to extract their data before pulling the plug on the service.…

'This was a deliberate, coordinated, digital attack'

Minnesota Governor Tim Walz has activated the state's National Guard and declared a state of emergency in response to a cyberattack on the city of Saint Paul.…

Journalist Jack Poulson accidentally discovered that Google had completely removed two of his articles from search results after someone exploited a vulnerability in the company's Refresh Outdated Content tool. The security flaw allowed malicious actors to de-list specific web pages by submitting URLs with altered capitalization to Google's recrawling system. When Google attempted to index these modified URLs, the system received 404 errors and subsequently removed all variations of the page from search results, including the original legitimate articles. The affected stories concerned tech CEO Delwin Maurice Blackman's 2021 arrest on felony domestic violence charges. In a statement to 404 Media, Google confirmed the vulnerability and said it had deployed a fix for the issue.

Read more of this story at Slashdot.

Can we have this as a global feature in all software? Please?

Zed, a fast new Rust-based text editor aimed at programmers, now lets you totally disable LLM bot integration. We're sure some users will rejoice – but how many?…

BrianFagioli writes: AI might be the future of software development, but a new report suggests we're not quite ready to take our hands off the wheel. Veracode has released its 2025 GenAI Code Security Report, and the findings are pretty alarming. Out of 80 carefully designed coding tasks completed by over 100 large language models, nearly 45 percent of the AI-generated code contained security flaws. That's not a small number. These are not minor bugs, either. We're talking about real vulnerabilities, with many falling under the OWASP Top 10, which highlights the most dangerous issues in modern web applications. The report found that when AI was given the option to write secure or insecure code, it picked the wrong path nearly half the time.

Read more of this story at Slashdot.

Competiton regulator smells abuse of a dominant market position, Zuckercorp claims all is well

Meta's addition of AI services to encrypted messaging platform WhatsApp has Italian officials suspecting the Silicon Valley giant may be abusing its dominant market position to push unwanted features on users.…

JPMorgan's proposed fees for customer data access would cost fintech startups between 60 and 100% of their annual revenue "just from one bank," according to a trade group representing the affected firms. Steve Boms, executive director of the Financial Data and Technology Association, said the charges would apply across all 30 companies in his group that received pricing notices from the nation's largest bank. The trade association, whose members include Plaid, Fiserv and Intuit, called JPMorgan's move a "pure and simple" attempt to kill competition that would "put third parties out of business altogether." The fees could take effect in September, ending more than a decade of free data access that fintech companies have used to build their business models. JPMorgan can now charge for data access after the Trump administration changed Consumer Financial Protection Bureau rules that previously prohibited such fees. The Financial Technology Association has taken the dispute to federal courts seeking to restore the Biden-era protections, while crypto trade groups have written directly to President Trump warning the fees would hurt digital currency companies.

Read more of this story at Slashdot.

The lure? Identity security and privileged access management tools to verify humans and... machines

Palo Alto Networks will buy Israeli security biz CyberArk in a $25 billion cash-and-stock deal confirmed today.…

Manager engagement has plummeted to its lowest level since tracking began, with only 27% of managers globally reporting they feel involved and enthusiastic about their work, according to Gallup's annual State of the Global Workplace report. The 3-percentage-point decline from 2023 marks an unprecedented drop in manager satisfaction. Overall employee engagement fell to 21% in 2024 from 23% the previous year, representing only the second decline in 15 years of data collection. The last drop occurred during 2020 COVID lockdowns. Female managers experienced the steepest decline at 7 percentage points, while younger managers fell 5 points. Managers now oversee nearly three times as many employees as in 2017, yet only 44% have received managerial training.

Read more of this story at Slashdot.

If at first you succeed, have some more money

NASA has awarded Firefly Aerospace $176.7 million to deliver a pair of rovers and a trio of scientific instruments to the Moon as part of the agency's Commercial Lunar Payload Services (CLPS) initiative.…

Brazil is the latest browser battlefield

Veteran browser maker Opera has filed a complaint with Brazil's Administrative Council for Economic Defense (CADE) against Microsoft over alleged anti-competitive practices in Windows that favor Edge.…

Mark Zuckerberg said Wednesday that Meta's AI systems have begun improving themselves over the past few months, calling the development "slow for now, but undeniable" and declaring that superintelligence is now within reach. The Meta CEO staked out the company's vision in a blog post for what he termed "personal superintelligence" -- AI that helps individuals achieve their goals rather than replacing human work entirely. Zuckerberg drew a sharp line between Meta's approach and that of other companies in the field, arguing that competitors want superintelligence "directed centrally towards automating all valuable work, and then humanity will live on a dole of its output." Meta's version would give people their own superintelligent assistants that know them deeply and help them create, experience adventures, and become better friends. Zuckerberg envisions smart glasses as the primary computing device, understanding context through what users see and hear throughout their day. The next few years represent a critical juncture, Zuckerberg wrote, calling the rest of this decade "the decisive period for determining the path this technology will take."

Read more of this story at Slashdot.

Distie insists global operations restored despite some websites only now coming back online

The cybercriminals claiming responsibility for Ingram Micro's ransomware attack put a deadline on leaking its data nearly a month after the raid.…

An anonymous reader quotes a report from CNBC: Google executives are pushing employees to act with more urgency in their use of artificial intelligence as the company looks for ways to cut costs. That was the message at an all-hands meeting last week, featuring CEO Sundar Pichai and Brian Saluzzo, who runs the teams building the technical foundation for Google's flagship products. "Anytime you go through a period of extraordinary investment, you respond by adding a lot of headcount, right?" Pichai said, according to audio obtained by CNBC. "But in this AI moment, I think we have to accomplish more by taking advantage of this transition to drive higher productivity. [...] We are competing with other companies in the world," Pichai said at the meeting. "There will be companies which will become more efficient through this moment in terms of employee productivity, which is why I think it's important to focus on that." [...] "We are going to be going through a period of much higher investment and I think we have to be frugal with our resources, and I would strive to be more productive and efficient as a company," Pichai said, adding that he's "very optimistic" about how Google is doing. At the meeting, Saluzzo highlighted a number of tools the company is building for software engineers, or SWEs, to help "everybody at Google be more AI-savvy." "We feel the urgency to really quickly and urgently get AI into more of the coding workflows to address top needs so you see a much more rapid increase in velocity," Saluzzo said. Saluzzo said Google has a portfolio of AI products available to employees "so folks can go faster." He mentioned an internal site called "AI Savvy Google" which has courses, toolkits and learning sessions, including some for individual product areas. Google's engineering education team, which develops courses for internal and external use, partnered with DeepMind on a training called "Building with Gemini" that the company will start promoting soon, Saluzzo said. He also referenced a new internal AI coding tool called Cider that helps software engineers with various aspects of the development process. Since May, when the company first introduced Cider, 50% of users tap the service on a weekly basis, Saluzzo said. Regarding Google's internal AI tools, Saluzzo said that employees should "expect them to continuously get better" and that "they'll become a pretty integral part of most SWE work."

Read more of this story at Slashdot.