Ximo F. Verde
LightNews LightNews
ximofverde.bsky.social
Ximo F. Verde
@ximofverde.bsky.social
23 followers 34 following 140 posts
Posts Replies Media Videos
ximofverde.bsky.social
How one controversial startup hopes to cool the planet ▶ And why many scientists are freaked out about the first serious for-profit company moving into the solar geoengineering field | MIT Technology Review www.technologyreview.com/2025/12/10/1...
But numerous researchers focused on solar geoengineering are deeply skeptical that Stardust will line up the government customers it would need to carry out a global deployment as early as 2035, the plan described in its earlier investor materials—and aghast at the suggestion that it ever expected to move that fast. They’re also highly critical of the idea that a company would take on the high-stakes task of setting the global temperature, rather than leaving it to publicly funded research programs.
December 11, 2025 at 9:27 AM
ximofverde.bsky.social
Large language mistake ▶ Cutting-edge research shows language is not the same as intelligence. The entire AI bubble is built on ignoring it | The Verge www.theverge.com/ai-artificia...
The AI hype machine relentlessly promotes the idea that we’re on the verge of creating something as intelligent as humans, or even “superintelligence” that will dwarf our own cognitive capacities. If we gather tons of data about the world, and combine this with ever more powerful computing power (read: Nvidia chips) to improve our statistical correlations, then presto, we’ll have AGI. Scaling is all we need. But this theory is seriously scientifically flawed. LLMs are simply tools that emulate the communicative function of language, not the separate and distinct cognitive process of thinking and reasoning, no matter how many data centers we build.
November 26, 2025 at 8:37 AM
ximofverde.bsky.social
Microsoft tries to head off the “novel security risks” of Windows 11 AI agents ▶ Agents with read/write access to your files create big security, privacy issues | Ars Technica arstechnica.com/gadgets/2025...
But these safeguards and monitoring capabilities don’t change the fact that you’re exposing yourself to privacy and security risks by using AI agents. They’ll be able to request read and write access to most of the files in your user account—by default, anything in the Documents, Downloads, Desktop, Music, Pictures, and Videos folders. They’ll have access to any apps that have been installed for all users on the PC (apps that have only been installed in your user account won’t be accessible to the agent, and it will also be possible for users to install apps that only their agents can access.) And agents can potentially be vulnerable to hijacking that exposes your data to attackers—Microsoft specifically mentions “cross-prompt injection (XPIA), where malicious content embedded in UI elements or documents can override agent instructions, leading to unintended actions like data exfiltration or malware installation.”
November 18, 2025 at 8:44 PM
ximofverde.bsky.social
LLMs show a “highly unreliable” capacity to describe their own internal processes ▶ Anthropic finds some LLM “self-awareness,” but “failures of introspection remain the norm.” | Ars Technica arstechnica.com/ai/2025/11/l...
The full paper on “Emergent Introspective Awareness in Large Language Models” uses some interesting methods to separate out the metaphorical “thought process” represented by an LLM’s artificial neurons from simple text output that purports to represent that process. In the end, though, the research finds that current AI models are “highly unreliable” at describing their own inner workings and that “failures of introspection remain the norm.”
November 4, 2025 at 8:31 AM
ximofverde.bsky.social
I tried OpenAI’s new Atlas browser but I still don’t know what it’s for ▶ My impression is that it is little more than cynicism masquerading as software | MIT Technology Review www.technologyreview.com/2025/10/27/1...
It’s been hard for me to understand why Atlas exists. Who is this browser for, exactly? Who is its customer? And the answer I have come to there is that Atlas is for OpenAI. The real customer, the true end user of Atlas, is not the person browsing websites, it is the company collecting data about what and how that person is browsing.
October 28, 2025 at 8:07 AM
ximofverde.bsky.social
Satellites Are Leaking the World’s Secrets: Calls, Texts, Military and Corporate Data ▶ With just $800 in basic equipment, researchers found a stunning variety of data sent by satellites unencrypted | Wired www.wired.com/story/satell...
For three years, the UCSD and UMD researchers developed and used an off-the-shelf, $800 satellite receiver system on the roof of a university building in the La Jolla seaside neighborhood of San Diego to pick up the communications of geosynchronous satellites in the small band of space visible from their Southern California vantage point. By simply pointing their dish at different satellites and spending months interpreting the obscure—but unprotected—signals they received from them, the researchers assembled an alarming collection of private data: They obtained samples of the contents of Americans’ calls and text messages on T-Mobile’s cellular network, data from airline passengers’ in-flight Wi-Fi browsing, communications to and from critical infrastructure such as electric utilities and offshore oil and gas platforms, and even US and Mexican military and law enforcement communications that revealed the locations of personnel, equipment, and facilities.
October 14, 2025 at 11:51 AM
ximofverde.bsky.social
The EU has a secret weapon to counter Trump’s economic bullying. It’s time to use it ▶ The anti-coercion instrument, or ‘trade bazooka’, is designed to shield against foreign pressure | The Guardian www.theguardian.com/commentisfre...
What is to be done? Europe’s anti-coercion instrument works by calculating the degree of the coercion and imposing counter-actions. Provided most European governments agree, the European Commission could kick US goods and services out of Europe’s market, or apply tariffs to them. It can strip their intellectual property rights, block their investments and require reparations as a condition of readmittance to Europe’s market.
October 9, 2025 at 7:13 AM
ximofverde.bsky.social
OpenAI’s Sora 2 Copyright Infringement Machine Features Nazi SpongeBobs and Criminal Pikachus ▶ The main use of Sora appears to generate brainrot of major beloved copyrighted characters + the millions of articles, images, and videos OpenAI has scraped | 404 Media www.404media.co/openais-sora...
The main immediate use of Sora so far appears to be to allow people to generate brainrot of major beloved copyrighted characters, to say nothing of the millions of articles, blogs, books, images, videos, photos, and pieces of art that OpenAI has scraped from people far less powerful than, say, Nintendo. As a reward for this wide scale theft, OpenAI gets a $500 billion valuation. And we get a tool that makes it even easier to flood the internet with slightly better looking bullshit at the low, low cost of nearly all of the intellectual property ever created by our species, the general concept of the nature of truth, the devaluation of art through an endless flooding of the zone, and the knock-on environmental, energy, and negative labor costs of this entire endeavor.
October 2, 2025 at 7:17 AM
ximofverde.bsky.social
Revealed: Europe losing 600 football pitches of nature and crop land a day ▶ Investigation shows extent of green land lost across UK and mainland Europe to development from 2018 to 2023 | The Guardian www.theguardian.com/environment/...
Nature accounts for the majority of the losses, at about 900 sq km a year, but the research shows we are also building on agricultural land at a rate of about 600 sq km a year, with grave consequences for the continent’s food security and health.
October 1, 2025 at 8:47 AM
ximofverde.bsky.social
Scientists make embryos from human skin DNA for first time ▶ The technique could overcome infertility due to old age or disease by using almost any cell in the body as the starting point for life | BBC News www.bbc.com/news/article...
October 1, 2025 at 7:35 AM
ximofverde.bsky.social
AI ‘Workslop’ Is Killing Productivity and Making Workers Miserable ▶ AI slop is taking over workplaces | 404 Media www.404media.co/ai-workslop-...
A joint study by Stanford University researchers and a workplace performance consulting firm published in the Harvard Business Review details the plight of workers who have to fix their colleagues’ AI-generated “workslop,” which they describe as work content that “masquerades as good work, but lacks the substance to meaningfully advance a given task.” The research, based on a survey of 1,150 workers, is the latest analysis to suggest that the injection of AI tools into the workplace has not resulted in some magic productivity boom and instead has just increased the amount of time that workers say they spend fixing low-quality AI-generated “work.”
September 26, 2025 at 9:03 AM
ximofverde.bsky.social
Trump is waging culture war on Europe by promoting rightwing allies, report finds ▶ Study says US president seeks to interfere in elections and ‘move the ideological centre of gravity’ in European politics | The Guardian www.theguardian.com/world/2025/s...
The study suggests conditions are “ripe” for European leaders to walk off Trump’s “film set”. Polling shows European sentiment – the sense of belonging to a common space, sharing a common future and subscribing to common values – is strong, it says. Eurobarometer data shows citizen trust in the EU is at its highest since 2007 and has risen in 12 countries – most sharply in Sweden, France, Denmark and Portugal – since Trump returned to the White House. Majorities in almost all feel attached to the EU. Hardly any party now wants to leave the EU, and majorities in all but three countries (Romania, Poland and the Czech Republic) say the role of the EU in protecting Europe’s citizens against global crises and security risks must become more important.
September 23, 2025 at 6:55 AM
ximofverde.bsky.social
The Software Engineers Paid to Fix Vibe Coded Messes ▶ LinkedIn has been joking about “vibe coding cleanup specialists,” but it’s actually a growing profession | 404 Media www.404media.co/the-software...
The alleged benefit of vibe coding, which refers to the practice of building software with AI-coding tools without much attention to the underlying code, is that it allows anyone to build a piece of software very quickly and easily. As we’ve previously reported, in reality, vibe coded projects could result in security issues or a recipe app that generates recipes for “Cyanide Ice Cream.” If the resulting software is so poor you need to hire a human specialist software engineer to come in and rewrite the vibe coded software, it defeats the entire purpose.
September 12, 2025 at 8:10 AM
ximofverde.bsky.social
When surgical tools don’t fit: how gender bias in design puts female surgeons at risk ▶ Built for men, used by everyone | The Conversation theconversation.com/when-surgica...
In vascular and cardiac surgery, precision and power go hand in hand. These procedures require surgeons to maintain awkward positions for extended periods, often in high-pressure situations. Even without design flaws, the risk of muscle and joint strain is significant. But when a handle is too big to grip securely, or a control requires more force than a surgeon can comfortably exert, that risk increases sharply and disproportionately for women.
September 10, 2025 at 5:28 AM
ximofverde.bsky.social
Massive Leak Shows How a Chinese Company Is Exporting the Great Firewall to the World ▶ Geedge Networks, a company w ties 2 the founder of China’s mass censorship infra, is selling its censorship & surveillance systems to ≥ 4 other countries in Asia and Africa | Wired www.wired.com/story/geedge...
Researchers who reviewed the leaked material found that the company is able to package advanced surveillance capabilities into what amounts to a commercialized version of the Great Firewall—a wholesale solution with both hardware that can be installed in any telecom data center and software operated by local government officers. The documents also discuss desired functions that the company is working on, such as cyberattack-for-hire and geofencing certain users.
September 9, 2025 at 2:21 PM
ximofverde.bsky.social
Former WhatsApp security boss in lawsuit likens Meta’s culture to a “cult” | ▶ Meta allegedly prioritized user growth over security, lawsuit said | Ars Technica arstechnica.com/security/202...
Shortly after assuming that role, the lawsuit said, Baig “discovered systemic cybersecurity failures that posed serious risks to user data.” During a red-team exercise designed to find and exploit security vulnerabilities so they can be fixed, Baig said he found that roughly 1,500 engineers inside the messenger division had “unrestricted access to user data, including personal information covered by the FTC Privacy Order, and could move or steal such data without detection or audit trail.”
September 8, 2025 at 8:43 PM
ximofverde.bsky.social
Five children see HIV viral loads vanish after taking antiretroviral drugs | ▶ The first widespread success in curing HIV may come from children, not adults | Ars Technica via Wired arstechnica.com/health/2025/...
August 3, 2025 at 12:28 PM
ximofverde.bsky.social
Google’s AI Is Destroying Search, the Internet, and Your Brain | ▶ Google’s AI Overview, which is easy to fool into stating nonsense as fact, is stopping people from finding and supporting small businesses and credible sources | 404 Media www.404media.co/googles-ai-i...
Essentially, the data shows that Google’s AI Overview feature introduced in 2023 replacing the “10 blue links” format that turned Google into the internet’s de facto traffic controller will end the flow of all that traffic almost completely and destroy the business of countless blogs and news sites in the process. Instead, Google will feed people into a faulty AI-powered alternative that is prone to errors it presents with so much confidence, we won’t even be able to tell that they are errors.
July 29, 2025 at 7:46 AM
ximofverde.bsky.social
The Media's Pivot to AI Is Not Real and Not Going to Work | ▶ AI is not going to save media companies, and forcing journalists to use AI is not a business model | FULL ARTICLE www.404media.co/the-medias-p...
So AI is destroying traffic, ripping off our work, creating slop that destroys discoverability and further undermines trust, and allows random people to create news-shaped objects that social media and search algorithms either can’t or don’t care to distinguish from real news. And yet media executives have decided that the only way to compete with this is to make their workers use AI to make content in a slightly more efficient way than they were already doing journalism.
July 16, 2025 at 11:13 AM
ximofverde.bsky.social
ChatGPT made up a product feature out of thin air, so this company created it | ▶ Soundslice caught OpenAI's bot telling users about a fake music notation feature - then built it | Ars Technica arstechnica.com/ai/2025/07/c...
On Monday, sheet music platform Soundslice says it developed a new feature after discovering that ChatGPT was incorrectly telling users the service could import ASCII tablature—a text-based guitar notation format the company had never supported. The incident reportedly marks what might be the first case of a business building functionality in direct response to an AI model's confabulation.
July 14, 2025 at 2:05 PM
ximofverde.bsky.social
‘I’ve £90k in student debt – for what?’ Graduates share their job-hunting woes amid the AI fallout | ▶ AI isn’t just taking away entry-levels jobs, it’s helping thousands apply for the same job with almost the same CV | The Guardian www.theguardian.com/money/2025/j...
Martyna, a 23-year-old who will receive a master’s degree in English literature from the University of York this autumn, was among other graduate jobseekers who got in touch with the Guardian via a callout and has been searching for her first full-time job since the beginning of May. “I’ve applied to about 150 entry-level jobs – in marketing, publishing, the civil service, charities, but also for retail and hospitality positions,” she said. “So far I’ve had five interviews, many almost instant rejections, plus ghosting. It makes me want to scream. “Platforms use AI to search for key words. I have friends who have copied entire job descriptions, pasted them into the Word document, reduced the font, and turned the colour to white so AIs find the words they’re looking for. It feels dystopian.”
July 13, 2025 at 9:32 AM
ximofverde.bsky.social
The Open-Source Software Saving the Internet From AI Bot Scrapers | ▶ Anubis, which block AI scrapers from scraping websites to death, has been downloaded almost 200,000 times | 404 Media www.404media.co/the-open-sou...
Essentially, Anubis verifies that any visitor to a site is a human using a browser as opposed to a bot. One of the ways it does this is by making the browser do a type of cryptographic math with JavaScript or other subtle checks that browsers do by default but bots have to be explicitly programmed to do. This check is invisible to the user, and most browsers since 2022 are able to complete this test. In theory, bot scrapers could pretend to be users with browsers as well, but the additional computational cost of doing so on the scale of scraping the entire internet would be huge. This way, Anubis creates a computational cost that is prohibitively expensive for AI scrapers that are hitting millions and millions of sites, but marginal for an individual user who is just using the internet like a human.
July 11, 2025 at 7:16 AM
ximofverde.bsky.social
AI won’t replace computer scientists any time soon – here are 10 reasons why | The Conversation theconversation.com/ai-wont-repl...
Where AI falls short What AI provides is more powerful tools for CS professionals to do their jobs. This means they can now take concepts further – from ideation to market deployment – while requiring fewer support roles and more technical leadership. There are, however, many areas where specialised human input is still essential, whether for trust, oversight or the need for human creativity. Examples abound, but there are 10 areas that stand out in particular: Adapting a hedge fund algorithm to new economic conditions. This requires algorithmic design and deep understanding of markets, not just reams of code. Diagnosing intermittent cloud service outages from providers like Google or Microsoft. AI can troubleshoot on a small scale, but it cannot contextualise large-scale, high-stakes troubleshooting. Rewriting code for quantum computers. AI cannot do this without extensive examples of successful implementations (which do not currently exist). Designing and securing a new cloud operating system. This involves high-level system architecture and rigorous testing that AI cannot perform. Creating energy-efficient AI systems. AI cannot spontaneously invent lower power GPU code, or reinvent its own architecture. Building secure, hacker-proof, real-time control software for nuclear power plants. This requires embedded systems expertise to be mixed with the translation of code and system design. Verifying that a surgical robot’s software works under unpredictable conditions. Safety-critical validation exceeds AI’s current scope. Designing systems to authenticate email sources and ensure integrity. This is a cryptographic and multi-disciplinary challenge. Auditing and improving AI-driven cancer prediction tools. This requires human oversight and continuous system validation. Building the next generation of safe and controllable AI. Evolving towards safer AI cannot be done by AI itself – this is a human responsibility.
July 1, 2025 at 12:29 PM
ximofverde.bsky.social
Anthropic releases custom AI chatbot for classified spy work | ▶ "Claude Gov" is already handling classified information for the US government. | Ars Techica arstechnica.com/ai/2025/06/a...
Of course, using AI models for intelligence analysis raises concerns about confabulation, where the models generate plausible-sounding, inaccurate information. Since AI neural networks operate on statistical probabilities rather than functioning as databases, they can potentially misinform intelligence agencies if not used properly. For example, the models may produce convincing but incorrect summaries or analyses of sensitive data, creating risks when accuracy is critical for national security decisions.
June 8, 2025 at 3:20 PM
ximofverde.bsky.social
| The real story isn’t young men supposedly voting far right. It’s what young women are up to | ▶ There is an opportunity staring centre-left parties in the face – if they reject the male gaze distorting our politics | The Guardian www.theguardian.com/commentisfre...
That the media chooses nevertheless to focus on young men illustrates the male gaze that continues to dominate society, which not only means that whatever men do or think is deemed inherently important and worthy of both academic and political attention, but also sets men as the norm. This ends up strengthening the far right’s political prospects: given that men are the norm, what they supposedly do or think is deemed “normal”, which means that all politicians should come up with “commonsense” policies to cater to them.
May 21, 2025 at 12:43 PM