Komunitas
feddit.org
Was auch blöd ist, dass die Open-Source Alternativen oft die suchterzeugenden Designelemente, wie endless scrolling, time lines, likes, von ihren “Vorbildern” kopieren. Dann hat man letztlich mit Mastodon und so weiter den Suchtstoff kopiert und säuft sich bildlich gespochen die Leber nun mit Schnaps aus dem Bioladen kaputt. Eine interessante Alternative zu den süchtig machenden Formaten a la Facebook und Twitter ist das Open Source Gemini-Netzwerkprotokoll, das trotz seines Namens nichts mit dem KI-Programm von Google zu tun hat (es entstand 2019/2020, bevor Google seine KI veröffentlichte). Mittlerweile hat sich ein Netzwerk mit einigen tausend dezentralen kleinen Servern gebildet, das organisch wächst. Das Gemini-Netzwerk ist, etwas verkürzt, eine Weiterentwicklung von gopher. Es handelt sich um eine extrem vereinfachte Software, die es einerseits erlaubt, textorientierte Inhalte sowie andere Medien in einem einfachen Browser (aber ohne merviges Javascript, Pop-Ups, Tracking, erzwungene Newsletter, Anzeigen, Logins, Cookie-Banner etc) anzusehen. Und andererseits extrem vereinfachte Server-Software, die es ermöglicht, seine eigenen Texte und Medien auf seinem eigenen kleinen Server zu hosten (ein stromsparender Raspberry Pi reicht dazu). Die unterstützten Formate sind Blogs, Mikroblogs, Journals (wie das alte LiveJournal), Webseiten wie Wordpress, beliebige Dateien und Medien (z.B. auch Download von Software). Es wird gerade hier auf c/de_EDV diskutiert: https://feddit.org/post/26001979
Komunitas
piefed.blahaj.zone
Differences to LibreWolf While the two projects have shared primary goals of privacy, security, and user freedom and Konform retains most of Librewolf’s patches, Konform takes a different stance on some questions and provides a different out-of-the-box experience compared to LibreWolf in name of the project goals. For example^1^: Based on Firefox ESR (Extended Support Release) instead of RR (Rapid Release) Less update churn and delayed access to newest features, while staying up to date with security fixes Removed integrations with online services for improved privacy and usability “RemoteSettings” is completely disabled by default, also disabling features depending on it Konform Browser will not download or sync updates of settings, preferences, search engines, or AI models at runtime All remote analytics and telemetry disabled “AI chatbot” doesn’t have proprietary providers (ChatGPT/Claude/Copilot/Gemini/) preconfigured Full-page machine translations work more reliably when offline Disable OCSP validation Upstream enforces OCSP validation by default, which means it will leak metadata to remote servers and not function in offline/airgapped environments Konform Browser instead enables more modern CRLite for certificate revocation checks Features Welcome screen (about:welcome) consisting of a preset switcher allowing user to choose between one of four default presets: Purely Private 🔒️ is ideal for private networks and when you need to keep external trust and communications to an absolute minimum. Disables all integrations with external servers and loading of opague binary blobs. Core Security 🛡️ is great when you want to keep external trust and communications to a minimum without compromising on security and common base features. Enables low-risk optional features and security-related updates from Mozilla (public suffix list, certificate revocation lists, and such). Basic Fuctionality✳️ is great for daily online surfing. Enables common optional features (like WebGPU and local ML) while keeping risky and potentially compromising integrations off. Just Make It Work🦊 is closest to common Firefox defaults. Unlocks RemoteSettings integrations to same extent as FF and re-enables potentially risky and leaky features depending on external providers for greatest out-of-box functionality and website compatibility. Useful for configuration testing and troubleshooting of website compatibility, and for non-sensitive scenarios where privacy is not a concern. Comes with Multi-Acccount Containers Lite pre-enabled for convenient tab compartmentalization Comes with bundled fonts matching Tor Browser and Mullvad Browser. Improved protection against font enumeration attacks while ensuring fonts render consistently regardless of what you have installed on system. about:translation and about:inference for direct access to translation and ML features Ported over bug- and security fixes from Tor Browser A couple of privacy-related patches not built elsewhere Earlier access to upstream security fixes Assorted privacy-strengthening of defaults Disable cross-origin referer by default Enable letterboxing by default if ResistFingerPrinting is enabled Extended configurability New preference privacy.resistFingerprinting.allowTheming (default true) Allows setting non-default theme when ResistFingerPrinting is enabled Dynamic light/dark theme depending on system preferences LibreWolf requires disabling ResistFingerPrinting for dark mode User-configurable FireFox Sync endpoints For using self-hosted or third-party Sync and Accounts servers User-configurable HTTP Referer spoofing Management UI for ML models and allow loading custom “AI” models from HuggingFace Link Preview feature configurable for use with local ML models Changes to RemoteSettings allowlists take effect without browser restart Reskinned logo and privacy-purple color scheme …etc Bundled extensions Librewolf will download uBlock Origin from Mozilla Addons on default profile initialization. Konform Browser does not download any extension by default. Instead, it will attempt to auto-detect and enable the following addons if already installed on the system by the user on a recognized path: uBlock Origin Alpine: ublock-origin Arch: firefox-ublock-origin Debian: webext-ublock-origin-firefox Fedora: mozilla-ublock-origin No-Script Arch: firefox-noscript Fedora: mozilla-noscript Privacy Badger Arch: firefox-extension-privacy-badger (AUR) Fedora: mozilla-privacy-badger This can be customized without rebuilding the browser by editing /usr/lib/konform/distribution/policies.json (https://codeberg.org/konform-browser/settings) to preload your own extensions or disable any of the defaults.
Komunitas
lemmy.sdf.org
The problem is that no browser can allow you to escape the horror that is web standards & practices that have been developed over decades […] practically the entire web is reliant on JavaScript, […] I’ve been saying it for a while: continuing to play catch is a losing move for Mozilla or for any independent browser maker. The real move, is to switch to or at least integrate an alternate internet, something that uses a protocol that is simpler and more limited by design - just get rid of Javascript (or of “remote execution”, really) and you instantly get a much leaner, much securer internet design. I’ve heard pretty good things about the Gemini protocol, but IMHO they went too far too extremist into the “text internet” philosophy, and as a result is a raw downgrade from Gopher. Gopher could actually be a good option.
Komunitas
midwest.social
“A beautiful size.” I like that. It describes the phenomenon exactly. Like, Gemini (the protocol) is too small. It’s below the threshold is usefulness. I still cross publish (serve) my content in Gemini format, but I never actually browse it anymore. (Tiny rant) That said, I think Gemini went too far; gmi is smaller than it needs to be, and it was locked down and made immutable prematurely. Client/server interactions need to be a little more complex; I think the extreme simplicity has contributed to its lack of adoption: it’s almost impossible to serve a functional and user-friendly search engine with, leading to what’s largely a dark network - not intentional dark, but filled with isolated, unreachable nodes and practically no discovery. And with its failure, it shut the door on that solution space. It was popular enough that early adapters who tried it are going to shy away from something Gemini-like, but a little more full-featured. I’m bitter about Gemini; if only Solderpunk had had a little more vision.
Komunitas
lemmy.ml
Oh completely, there will be blood not global extinction though. It’s accelerationism. Musk removed the government bodies investigating the safety of neuralink, as soon as that happened Meta announced their own brain chip. That’s why Joe Rogan, close personal friend of Elon Musk, has been used to plant ideas of technology getting into your body. Remember him talking about how cool it’ll be? Nobody will want it “Until you’re flying around the neighbourhood, shooting lazers out of your eyes”. Big Tech all has one goal, Ads beamed directly into your brain. Permanent “content.” Unending novelty to quell any uprising, that’s why starlink exists. Never out of range of wifi and that is what generative AI is for. You remember the chip shortage? Why it’s a thousand dollars now for a half decent nvidia card? It’s because Generative AI lives and dies by the number of threads it can process at once and that is dictated by the GPU array. That’s why the chip shortage started during covid and then afterwards Microsoft, Google, Brave, Duck Duck Go. Everyone has an AI Assistant. The embedded AI in these browsers and on your phone is a smaller version of the one running on their servers. It can see everything on your phone, always and report back.We know this already, when apple announced that they had identified users sharing child pornography and that their phones AI had identified them. Ignoring the obvious moral question of “how did they train their AI’s to recognise child sex abuse material?” That’s a statement that you have no privacy in anything you do with these products anymore. There is no end-to-end encryption that can protect you from that. That’s why LLM were created, 1st for surveillance, 2nd to be able to search the ungodly amount of data that Google and Apple and Microsoft have collected on all of us over the last quarter century. Using commonalities in search patterns and browser telemetry to track all of your movements online. To proliferate, wifi cameras all over the world, replacing the bulk of traditional closed circuit television systems. Instead we have computers with cameras, that run android. The automatic checkouts that ballooned in size and quantity during the pandemic, reducing the number of workers required and recording your grocery list, card number and taking your picture when you check out. Just like your clock in machine at work that runs on a tablet that runs android and takes your picture. Why does the clock in machine need to record you? Why Does alexa and google home always need to be listening to you? Curtis Yarvin is a self-titled extremist and technocrat. One who believes that society should be run by enlightened scientists (See: delusional tech bros who cannot relate to normal people because their actions and lives are no longer dictated by the drive to accrue capital). Yarvin said in 2008 that the goal of a technocratic society should be to “Virtualise, non-productive people as a humane alternative to genocide.” They want to plug us in, and allow those they have no use for to live out the rest of their lives virtually as they slowly die from dehydration and starvation over a few weeks. Those they find useful will be forced to work to rent their homes in newly developed tech cities. There will be no physical cash. All transactions will be recorded and monitored any unique idea will be recorded to prevent anyone else from accruing the capital to “ascend” to their class. Dissent will be monitored through AI, Hey Grok/Gemini/ Copilot etc. Is anyone voicing distrust or unhappiness with the regime? This sounds absurd, I know. But I must point out, Curtis Yarvin is regularly around the whitehouse, he is also a close personal friend of Peter Thiel and mentor to JD Vance. That is where the far-right came from, it came from Big Tech. It created a vacuum of entry level jobs absorbed by Uber, Lyft, Skip the dishes (Just Eat) allowing for undocumented migrants to use the app to usurp food delivery and taxi services. Skirt labour laws and remove the need to supply insurance and pay minimum wage. It created a job shortage, leading to higher rates of competition and fewer working young men. Social Media and the entire clearnet, has been centralised to a point of absolutionist. control. They algorithmically feed us all of our information, through the phone, they divided us, so there is no discourse. They created this cognitive dissonance, where we have been conditioned to disregard IMMEDIATELY, other peoples ideas. We are shown information we already believe to be correct, under the illusion that this is what everyone sees. The internet is nothing but a giant skinner box, moulded to create these conditions. That’s why enlightened scientist, champion of logic and reason, Jordan B. Peterson, suddenly does not understand global warming. The science we explain to 5 year olds. There is no culture war, there is just a generation of radicalised men, manipulated into war by billionaires to finally go mask off and seize control of all of Earth’s resources. They want to speed global warming, strip mine the entire planet, while we are huddled in their earthscrapers and techno cities, and then fuck off to Mars. They have talked openly about this for the last 25 years at least, it is not a secret. The first earth scraper has already been approved for construction. In Hungary, you know, that place with Orban. Trumps authoritarian buddy. They need fascism, they need intolerance to be allowed to push us into global nuclear conflict. If all this sounds impossible, I’d just like to point out, that the former director of the Technocratic Movement is Elon Musk’s Maternal grandfather. He was an absolute fraud, like Ellen and Chiropractor. They aren’t enlightened scientists. They are racists, global terrorists and the single largest organised spy network the world has ever known. We need to burn them -the fuck- down. Smart Tech is to make you stupid. Autocorrect, so you can’t spell without it. Google maps to track everywhere you go and keep you looking at a screen so you don’t commit local scenery to memory and you can’t find your way out of a paper bag without them knowing. Every person who has ever used any of their voice recording software, has been voice printed. I mean, maybe it wasn’t a good idea to let 5 companies map every single above ground man made structure on Earth, every single road system and come, preinstalled. Monitoring every single thing you have ever logged into a search bar. Every automatic backup you did not know was happening because they made it opt out. Every passport scan, credit card number, debit card number license plate. Every persons medical history, finger prints, home address and social circle.
Komunitas
lemmy.world
The Gemini podcast is going to condense your text and make it conversational, but it will necessarily lose detail in the process. A better recommendation is the Eleven Labs Reader, it’ll just read any text or file you throw at it with top tier voice models. Can use it for free and they have paid plans for more use. They also have a “podcast” generator option like Gemini, but I haven’t tried it so can’t vouch for the quality. I use Eleven Labs all the time for things I want to read, like email newsletters, industry publications, etc but never find the time to sit down and read. Now I can have AI read them to me while I walk the dog. Super handy imo
Komunitas
feddit.de
Mhm I disagree with your second point. Since you can’t use any styling on Gemini objects, you won’t get table layout as we had in dark ages of Html. With tables like in Markdown you can just lay out tabular data in an actual table. Mhm I guess with the plaintext environment we still can link to external resources like images and other multimedia or interactives.
Komunitas
lemmy.world
While true, some things we want to simplify are sometimes as simple as they can be. But saying that, I’m thinking of Java, not of the Web. Java is really a wonderful creation. Just - sometimes when learning new things I understand that yes, I was right and some thing is too complex and it’s just that, but sometimes that it’s optimal and the “simple” way is even more complex. IMHO the Web solves two goals, which should be separated. Global hypertext services and serving applications executed on client in a sandbox. The latter is far more complex and demanding for security and efficiency and features, but the former is far more important socially. Maybe the former should rely upon a simpler and easier technology, like Gemini, and the latter be a kind of applications like an address book or a 2FA application. Where you see a list of imported connections, press “run”, and then over a standard protocol fetch the actual executable application to run in a sandbox. What the Web in practice already is for most people, untied from a global hypertext system. So that we’d have both. I mean, it’s pretty normal to open magnet links in a different application, or download an RDP connection file and open it in an RDP client. OK, my brain is asleep.
Komunitas
discuss.online
From the About page: This website is a proxy service that exposes gemini:// servers (and other hipster protocols, collectively referred to as the “smolnet”) over the web so they can be accessed in a normal web browser. And if you’re wondering about what gemini:// is, Wikipedia has an article, but it’s basically HTTP/HTML/JS/CSS stripped down to the absolute basics as an attempt to get rid of all the BS that has accumulated due to the corporate influence over the years.
Komunitas
lemmy.sdf.org
Archived In early 2025, the Chinese company DeepSeek launched a powerful LLM-based chatbot that quickly drew international attention. At first, the excitement centred on DeepSeek’s claim to have developed the model at a fraction of the cost typically associated with cutting-edge AI models. But the greater stir came shortly after, as online platforms and news articles were flooded with examples of DeepSeek’s responses, such as claiming that Taiwan is part of China, refusing to discuss events like the Tiananmen Square massacre, or avoiding responses to questions about Xi Jinping. […] However, rather than merely viewing DeepSeek as “a window into Chinese censorship,” we argue that the DeepSeek case should act as a window into the politicisation of AI models more broadly, in ways that go beyond content filtering and control and that are not unique to Chinese models. Of Course It’s Censored The fact that DeepSeek filters out politically sensitive responses is hardly surprising. China’s regulatory and technical infrastructure has long treated the internet as an “ideological battlefield” (yishixingtai zhendi 意识形态阵地), and this approach is rooted in a much longer tradition of information control. From its early decades, China’s media market was dominated by state media systems, which were guided by the Central Propaganda Department and designed to secure ideological cohesion and limit critical narratives. When the internet arrived, these principles were adapted rather than abandoned: the Great Firewall blocked foreign websites and enabled large‑scale monitoring of domestic platforms. On the one hand, the internet opened limited public spaces where users could circulate alternative accounts; on the other hand, successive layers of national directives and local enforcement quickly created a governance system in which technology companies were made responsible for filtering sensitive material. Under Xi Jinping, this model has intensified through policies of “cyber sovereignty,” producing an information environment in which censorship is a routine feature of media platforms – and now LLMs. […] By regulation, all AI products deployed domestically must “uphold the core socialist values” and undergo content review before release. Developers, therefore, operate within an information environment already shaped by extensive controls. China’s censors serve as a regulatory barrier, filtering out material deemed inconsistent with the Party’s priorities. In practice, this means that (1) the local training data available to developers is already censored, as certain content is largely absent from domestic news, search engines, and social media; (2) the model‑building process itself is conducted under compliance requirements; and (3) real‑time mechanisms are embedded, ensuring that certain prompts trigger avoidance scripts or canned replies. […] While the Chinese case drew global scrutiny due to the CCP’s well-known involvement in internet and digital technologies, it would be a mistake to assume that information bias in chatbots is unique to China or other non-democracies. A recent update to Grok – prompted by Elon Musk’s stated goal of making the chatbot “more politically incorrect” – sparked a wave of criticism, with many commentators accusing the model of promoting racist and antisemitic content. Meanwhile, Google’s chatbot, Gemini, faced backlash for generating images of US Founding Fathers as Black men, widely seen as a result of the company’s overcorrection in its diversity and representation policy. If so, these models, too, are biased. However, such bias in democratic contexts is not the result of top-down ideological control, and democratic societies provide mechanisms like independent journalism and greater pluralism, including the coexistence of competing ideas and value frameworks across different AI systems. […] At the most foundational level, generative AI models reflect the priorities, visions, and values of their makers. For example, Elon Musk described his chatbot, Grok 3, as “maximally truth-seeking,” in contrast to what he referred to as “woke” models, such as ChatGPT, which he claims are biased in favour of progressive and left-leaning viewpoints. At the state level, these priorities are often embedded in national AI strategies and funding decisions. Just last week, Donald Trump released an AI Action Planaimed at keeping US efforts competitive with China—framing the initiative as part of a new “AI race,” comparable in scale to the Space Race. Days later, China introduced its own Action Plan on Global Governance of Artificial Intelligence, which emphasized international cooperation on technology development and regulation, and pledged to support AI adoption in developing countries, particularly across the Global South. […] Conclusion Focusing narrowly on output censorship misses the forest for the trees. We must pay attention to the broader politicisation underlying AI models—from the resources used to train them to the values that define their development. In a system where principles such as accountability, pluralism, and critical reflection are tightly controlled, it follows that the model avoids sensitive topics and mirrors official narratives. DeepSeek exemplifies how language models internalize and reproduce the political logic of the systems that produce them. Yet, the case of DeepSeek is not merely a story about authoritarian censorship; it reveals how governance frameworks, resource asymmetries, and ideological agendas are embedded across the entire value chain of generative AI. […] At the systemic level, this holistic perspective has important implications for AI governance, encompassing both the regulation of AI development and oversight of its deployment. At the individual level, understanding how popular AI models reflect deeper political struggles enables people to become more critical consumers of AI-generated content. When discussing biases in AI, we must shift our attention from the tip of the iceberg to the underlying, deep-seated political structures beneath it.
Komunitas
ibbit.at
Aside from a few stand-out programs — looking at you, Star Trek — by the late 1960s, TV had already become the “vast wasteland” predicted almost a decade earlier by Newton Minnow. But for the technically inclined, the period offered no end of engaging content in the form of wall-to-wall coverage of anything and everything to do with the run-up to the Apollo moon landings. It was the best thing on TV, and even the endless press conferences beat watching a rerun of Gilligan’s Island. At the time, most of the attention landed on the manned missions, with the photogenic and courageous astronauts of the Mercury, Gemini, and Apollo programs very much in the limelight. But for our money, it was the unmanned missions where the real heroics were on display, starring the less-photogenic but arguably vastly more important engineers and scientists who made it all possible. It probably didn’t do much for the general public, but it sure inspired a generation of future scientists and engineers. With that in mind, we were pleased to see this Surveyor 1 documentary from Retro Space HD pop up in our feed the other day. It appears to be a compilation of news coverage and documentaries about the mission, which took place in the summer of 1966 and became the first lunar lander to set down softly on the Moon’s surface. The rationale of the mission boiled down to one simple fact: we had no idea what the properties of the lunar surface were. The Surveyor program was designed to take the lay of the land, and Surveyor 1 in particular was tasked with exploring the mechanical properties of the lunar regolith, primarily to make sure that the Apollo astronauts wouldn’t be swallowed whole when they eventually made the trip President Kennedy had mandated back in 1961. The video below really captures the spirit of these early missions, a time when there were far more unknowns than knowns, and disaster always seemed to be right around the corner. Even the launch system for Surveyor, the Atlas-Centaur booster, was a wild card, having only recently emerged from an accelerated testing program that was rife with spectacular failures. The other thing the film captures well is the spacecraft’s nail-biting descent and landing, attended not only by the short-sleeved and skinny-tied engineers but by a large number of obvious civilians, including a few lucky children. They were all there to witness history and see the first grainy but glorious pictures from the Moon, captured by a craft that seemed to have only just barely gotten there in one piece. The film is loaded with vintage tech gems, of course, along with classic examples of the animations used at the time to illustrate the abstract concepts of spaceflight to the general public. These sequences really bring back the excitement of the time, at least for those of us whose imaginations were captured by the space program and the deeds of these nervous men and women. NASA wants to return to the moon. They also want you to help. Turns out making a good landing on the moon is harder than you might think. From Blog – Hackaday via this RSS feed
Komunitas
ibbit.at
European Commission President Ursula von der Leyen and US President Donald Trump at Turnberry Hotel and Resort. Photo: European Commission Donald Trump is a busy man. He’s invading his country’s capital city, taking over cultural institutions, restricting democratic rights, running an influence campaign inside a supposed allied nation, and seemingly preparing to attack Venezuela as he continues to enable Israel’s genocide in Gaza — and that’s just the tip of the iceberg. On Tuesday, he must have received a reminder of whose money paved his way back to the White House. “I will stand up to Countries that attack our incredible American Tech Companies,” he declared from his Truth Social account. “With this TRUTH, I put all Countries with Digital Taxes, Legislation, Rules, or Regulations, on notice that unless these discriminatory actions are removed, I, as President of the United States, will impose substantial additional Tariffs on that Country’s Exports to the U.S.A., and institute Export restrictions on our Highly Protected Technology and Chips.” Despite the recent wave of supposed deals — most of which aren’t even written down, showing how truly useless they are — his declaration of (economic) war against any country trying to rein in US tech companies or wrest back some semblance of control over their tech sectors should put to bed any notion the intense period of trade hostilities is over. Trump is continuing to wield what might the US has left, and will do so for as long as he has it. The United States is a rogue nation spurred on by rogue companies in Silicon Valley and beyond that are intent on using the power at their disposal to extract maximum short-term benefit regardless of the long-term cost. It’s about time other countries give up their appeasement campaign and get serious about isolating the declining hegemon and its tech oligarchy until they’re forced to play nice. Disconnect is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. Bootlicking world leaders In my own country of Canada, we’ve been treated to months of concessions to the tech industry and the US government with little to show for it. Previous efforts to regulate AI have been thrown in the bin, as the new AI minister brags about using Google Gemini to make podcasts about legislation he’s supposed to understand about his job. The Justice Minister announced rules meant to protect young people online are also effectively dead, while Prime Minister Mark Carney has killed a capital gains tax increase that angered tech CEOs then rescinded the digital services tax to try to keep US officials at the negotiating table. Trump levied 35% tariffs anyway, then moved on to targeting lumber imports. Carney’s actions are cowardly in their own right, but no image better sums up the appeasement campaign of Western allies of the United States than European Commission President Ursula von der Leyen sitting next to Trump at his Scottish golf course, trying to justify accepting a terrible trade framework. It was at best marginally better than taking no deal at all, and even then the two sides started disputing what the actual terms of the deal were only days later. Some European officials felt disagreements over its tech regulations were settled. Trump and US tech CEOs clearly had other ideas. As one European diplomat put it, “Concessions are seen by him as a sign of weakness making him come back for more.” These are just a few examples of the cowardice put on display by Western leaders in recent months. As Trump threatens their sovereignty, seeks to destroy key sectors of their economies, and tries to extract even greater gains for the United States regardless of the cost to supposed allies, these supposed leaders have barely been able to mutter criticism of his actions toward them — let alone his belligerence toward other nations and the crackdown on his own citizens. They have become adept at appeasement, and it’s about time they break out of it. The US president will not change, regardless of the pleasantries they use when they greet him or the investment commitments they make to keep him happy. There might be pain to breaking with the United States or refusing to fold in the face of demands from the White House. But how far can they let this go while maintaining some degree of legitimacy and being able to look their own citizens in the eyes as they watch their leaders shy away from threats, bullying, and attacks on their countries? In Canada, Carney came to power promising “elbows up” against the United States. The opposition leader recently noted his “elbows have mysteriously gone missing.” Tech dependence empowers the US Trump’s broadside on tech policy should be an opportunity for other Western governments to end this embarrassing charade and show not only some self-respect, but that they have other options available to them. The global reach of US tech companies gives them immense power over governments, and grants them advantages that no company without such scale can match. Their billions of users improve their products and give them more data than most companies could ever imagine, and that becomes a clear competitive advantage. On top of that, every time foreign governments, companies, and users pay for or even use the services that US tech companies offer, they are creating value that ultimately flows back to the United States. In recent years, there has been a lot of focus on why US GDP per capita was growing faster than its peers or why the US stock market was comparatively performing so well. The truth is that having many of the world’s dominant tech companies, which also happen to be some of the most valuable corporations on the planet, headquartered there makes a big difference. That advantage granted to the United States by our collective dependence on US tech companies will not change unless we ween ourselves off their products and services. We certainly cannot allow generative AI to be used to entrench and expand the extent of that dependence, which some critics are even comparing to a colonial relationship. The only way to challenge the tech industry’s power is to challenge their global scale, and that means ramping up regulatory efforts and even forcing them out of their most important non-US markets. Appeasing Donald Trump and the Silicon elite will never deliver that outcome. All we need to do is look at India to see the futility of making concessions to a US regime that will only keep demanding more. Modi and Trump were assumed to have a good relationship, the country made trade concessions to keep the White House happy, and it killed its 6% digital ad tax in March in response to US trade concerns. Yet that didn’t save the country from the imposition of a 50% tariff rate — among the highest in the world. Now India is openly talking about self-reliance and getting closer to China. Instead of sleepwalking into a “century of humiliation,” as some commentators suggest Europe could be heading toward because it’s so yoked to the United States, it’s time for countries to defend their sovereignty and spurn US demands, even if it comes with some short-term pain. This is a time to be bold. Share Cooperation is essential New alliances are possible in this moment. Non-US Western countries are already building stronger defense alliances. The European Union is looking at forging closer ties to members of the CPTPP, which includes Canada, Australia, Japan, and other Pacific nations. Members of the BRICS are also making a play to wrench back some of their sovereignty from a world order that privileged the US and, to a lesser degree, other Western states. The more countries that shift from dependence on fossil fuels to electrification will also reduce the influence of the United States. At this crucial moment, the European Union should reach out its arms to traditional allies like Canada, Australia, Japan, and South Korea rather than turning inward, not to mention working with a broader grouping of countries like Brazil, Chile, and South Africa (just to name a few) to develop a common front against US aggression and, in particular, the colonial nature of its tech monopolists. It’s long past time to implement sweeping restrictions on data collection and transfer, stronger labor protections to end the push to precarity enabled by digital platforms, and stringent regulations targeting the harms of the tech infrastructure developed over the past several decades, including everything from social media to the pervasive surveillance culture that has come along with the business model of Silicon Valley. Earlier this year, the European Union was floating a much more aggressive retaliatory policy toward the United States. Whereas the trade war is hitting goods, EU officials were explicitly looking at options to target trade in services, with a specific focus on the immense quantity of services it contracts from US tech companies. For companies trying to use the Trump administration to get rid of foreign taxes and regulations, it would have been a major blow. But the European Union backed down and other governments have not tried something similar. They didn’t want to provoke the ire of the White House and Silicon Valley — but they should do just that. If countries want to escape the belligerence of the United States and its tech industry, they need to form stronger alliances and stop allowing the Trump administration to pick them off individually. Defense cooperation should be a stepping stone to deeper and broader alliances focused on tech development. Europe, Canada, and other countries may feel limited because they’ve allowed their security to become dependent on the United States, but even if they build up their militaries to claw that back, they’ll find that if they’re still dependent on US tech, they won’t have regained much real authority. We must stop US tech dystopia For decades, Silicon Valley has put growth, profits, and the expansion of its power ahead of all else. Those goals used to be obscured behind effective public relations campaigns to make people believe they were building the future and would not “be evil,” but that was only a strategy to displace incumbents and ultimately take their place. For the past decade, the drawbacks of the model they built have become increasingly apparent, but governments restrained themselves from properly addressing them for fear of appearing to spurn “innovation” and scare away precious tech investment dollars. Today, as the United States lashes out at friend even more than foe, the harms of the tech reality we’ve been made to depend on are impossible to ignore. Private companies have built out the most comprehensive surveillance apparatus in human history. Social media platforms are designed to keep us tethered to screens, fearful of the world beyond, and feeling terrible about ourselves. Gig apps turn us into algorithmically controlled wage slaves with no power over our work. Now generative AI has entered the picture to fill our feeds with engagement slop and convince us we should confide in chatbots instead of building real relationships — even as the cases of “AI psychosis” and AI-enabled suicide grow by the day. This cannot continue. Governments have a responsibility to their citizens to end the madness the tech industry has unleashed on us through strict and aggressive regulatory efforts, while getting serious about a form of digital sovereignty that builds an alternative much more focused on public benefit than maximizing shareholder value. Any campaign seeking to achieve those goals will come into conflict with the US government and the titans of Silicon Valley. There couldn’t be a better time to pick that fight and find new allies in the process. Disconnect is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. From Disconnect via this RSS feed
Komunitas
lemmygrad.ml
A lecturer in computer science and a polytechnician by training, Dr. Olivier Lompo is a lecturer and researcher at the University of Franche-Comté in France and at the Swiss Federal Institute of Technology in Lausanne in Switzerland. He teaches modules such as Data Mining, Natural Language Processing (NLP), Computer Vision, Internet of Things (IoT) and Wireless Sensor Networks, Artificial Intelligence (AI) for Cyber-Physical Systems, Cybersecurity and AI. His research focuses on the intelligent optimization of resource allocation and data transmission in wireless sensor networks using distributed AI. In this interview, he takes stock of the state of AI in the world and discusses the issues and challenges related to the development of AI in Africa. Through the Burkina Faso AI House, which he launched in July 2025, Dr. Olivier Lompo aims to help promote the use of AI in key sectors such as health, agriculture, education, cybersecurity, and cyberdefense. Sidwaya (S): Today, AI is revolutionizing all sectors of activity. What role does AI play in the socioeconomic development of the world today? Dr. Olivier Lompo (OL): It’s important to note that AI is reshaping the structures of our society, and its influence extends far beyond simple automation. AI is profoundly transforming working methods and economic models. For example, when you look at human and animal health, we can perform early diagnoses, personalized medicine, and even surgery assisted by companion robots. Modeling the 3-dimensional structure of a cell was made possible thanks to AI and clearly through the DINO v2 process. Genomics and gene therapies, stem cells, nanomedicine, restorative nanotechnologies, hybridization between man and machine are all technologies that will revolutionize all our relationships with the world in a few generations. It is also likely that life expectancy will increase at least during the 21st century. That’s great! You see, in the field of finance, fraud detection, automated risk management is a major advance today, it may be trivial but ask yourself, if this were not controlled, what would have become of the colossal masses of financial data to be analyzed? When discussing the role of AI in global economic development, it is important to remember that it has become a huge strategic lever for growth, competitiveness, and especially innovation. When you look at the countries that have adopted AI and made it their strategic ally, you will notice that the automation of repetitive tasks has freed up time for higher value-added activities. We now know that thanks to AI, banking services can reach unbanked populations in the case of African countries. Social and economic gaps could be quickly observed and accentuated, especially by new professions in Tech. Other major advances we are experiencing are also in energy optimization, given the importance of energy today, mastery of climate forecasts, and most importantly, mastery in the management of our agricultural resources. States that quickly integrate AI into their public policies are also taking a step ahead in competitiveness and social performance. AI is no longer an option; it has become an essential lever for growth and sovereignty. S: What is your assessment of the development of AI in Africa, in general, and in Burkina Faso, in particular? OL: It should be noted that Africa is already in the process of revolution and has been involved for a decade in developing and deploying some Artificial Intelligence solutions for its population. What we must remember is the scope of African AI solutions; on the continent, we tend to think locally, to develop IT projects with a continental impact. In my opinion, I hope that AI engineers today can think a little more globally, globally, without straying from African AI engineering. We have, for example, Copyly AI or even NextGit AI, to name just a few, which are African AIs developed by African engineers. It should be noted that we had more than a thousand African AI tools, but until then with little social or economic impact. In Burkina Faso, things are bubbling! Many AI projects, innovative solutions are being created. The government’s effort in this direction is very beneficial when you look at the various AI research laboratories like LAMIBIO and CITADEL, we understand how much doctoral students and researchers are advancing AI projects. We also have the ecosystem of academics who design, develop, and test AI tools locally. S: Can AI be an alternative to help Africa make up for its development delays in many areas? OL: I would say a resounding “YES,” considering the exciting projects underway on the continent. We know that AI can compensate for the lack of skilled human resources in certain areas, such as remote medical diagnosis, machine translation for African languages, and chatbots for public and banking services. This could enable governments and businesses to offer more services with less, by reducing associated costs and expanding access. I take as an example the possibility of developing a tool to identify students with learning difficulties with the possibility of early detection of school dropouts. AI can also help in the fight against corruption for good governance. With the current security situation, it is a good ally that must be pampered, AI can be used to develop conflict prevention and even prediction software. This will clearly help in the control of conflicts and social and economic crises. Africa can build its own AI model: ethical, frugal, inclusive, and rooted in its realities. We have Mistral and Claude with Europe, Open AI, Meta AI, or Gemini with the United States, Deep Seek with China, and where does Africa fit in? A revolutionary AI is coming in September 2025, and it’s the one we developed at the AI Research Center of the Swiss Federal Institute of Technology in Lausanne and the ETH Zurich in Switzerland. It’s a 100% Swiss, free language model with true openness to data. S: AI doesn’t just have its positive side; it also has its downside. With the advent of AI, many fear massive job losses, not to mention the ethical or moral questions that could arise from its use. As an expert on the subject, what’s the reality? Is there a danger ahead? OL: These fears are legitimate, but, in my opinion, misguided. AI could certainly eliminate jobs, but it transforms them profoundly, and it is in this process of change that new professions and jobs are created. We must therefore anticipate this change by training young people and workers in the skills of tomorrow. As for ethics, it is at the heart of our approach. We must define AI governance rooted in our African values, avoid algorithmic biases by opening up data, guarantee data protection through the Commission on Information Technology and Civil Liberties, and promote trustworthy AI. This is also the role of the Maison de l’AI: to train, supervise, and regulate. S: Do you feel that in our tropics, there is a clear awareness of the importance of AI as an essential vector of change and socio-economic development on the continent? OL: Awareness is emerging, but it remains elitist and concentrated in a few circles. We need to popularize, democratize, and raise awareness among decision-makers, young people, and even rural people. AI must be seen as a development tool. Through our upcoming programs starting this August, and our publications on LinkedIn, Facebook, and Instagram, the Maison de l’AI is committed to making AI accessible to all Burkinabés. S: What should Africa, Burkina Faso, do to make the most of AI in order to effectively respond to development challenges? OL: Invest massively in training, create innovation hubs, promote and make available local data, promote technological entrepreneurship and, above all, establish a clear and effective regulatory framework. Burkina Faso must notably integrate AI into its national development strategy by creating a specially dedicated Artificial Intelligence and Digital Agency, promote applied research and encourage public-private partnerships. Q: Burkina Faso is facing a security crisis. How can AI solutions help this country effectively fight the terrorist hydra? OL: AI can strengthen territorial surveillance through image recognition (drones), analyze weak signals on social networks, anticipate suspicious movements using predictive models, and optimize military logistics. It can also facilitate the management of internally displaced persons, the prevention of radicalization, and the securing of sensitive sites. However, these solutions must be developed locally and respect our fundamental jurisdictions. It should be noted that the Maison de l’AI is currently working in partnership with ANSSI, BCLCC, DCSSIC, DGTI, and soon with CIL, to reflect on, train and raise awareness, develop, and deploy AI tools that could clearly help to gradually control insecurity linked to terrorism and internal security. Q: As an AI teacher-researcher, what are your research areas or topics? What is the current state of AI research worldwide? OL: AI research is dominated by major technological powers like the United States, which has the largest universities in the world, including Stanford, MIT,Berkeley, which leads fundamental research. We also know that Google, Apple, Meta, Amazon, and Microsoft dominate language models with GPT, Gemini, and Claude, to name a few. Massive investments in deep learning, self-supervised learning, LLMs and LAMs, robotics, and embedded AI are being observed. China, for its part, remains the accelerated rise with massive state support to become a world leader by 2030. Leaders such as Alibaba, Tencent, and Baidu are developing powerful models like Ernie and ChatGLM. Concretely, we are seeing a breakthrough in research that is now very advanced in computer vision, military AI, surveillance, and predictive modeling. Europe is carrying out ethics-focused initiatives such as AI ACT, for which I worked within the European Commission with my research department FEMTO-ST. We are seeing a strong presence of European AI in cognitive robotics, digital health and much more in fundamental research. Some AI centers such as INRIA in France, TUM in Germany, or EPFL in Switzerland have very high-level productions. It should be noted that Canada also has a very high-level production through the Yoshua Bengio School, the same one who received the Nobel Prize in Physics in October 2024 with his MILA laboratory, a pioneer in responsible AI and fundamental AI. Regarding Africa, we can notice the emergence of some research centers like CITADEL, AI4D, DSN. It should be noted that I work on the intelligent optimization of resource allocation and data transmission in wireless sensor networks using distributed artificial intelligence. This is an area where, within the FEMTO-ST laboratory, my team and I were able to work on how AI algorithms can be integrated into sensor nodes to improve energy efficiency while reducing communication latency. The whole point is to successfully optimize transmission routes in wireless communication systems. Let’s start with this concrete example: Imagine a 20-hectare farm equipped with LoRa sensors and connected weather stations. The system will predict a local drought 48 hours in advance, automatically activate irrigation in critical areas, and finally prioritize sending alerts to high-yield areas. There are just as many areas of application. Q: What are the latest developments in AI or the latest research topics in AI? OL: I would say that the trends today are for the foundation models GPT-4.5, Claude 3, Gemini Ultra, LLaMA 3 towards a convergence between NLP, multimodality, and artificial cognition. With generative AI, we can clearly succeed in creating text, images, sound, video, especially with the JEPA model, which is Joint Embedding Predictive Architecture. As for self-supervised learning, we will see the gradual replacement of supervised learning in many fields such as NLP. And the very latest tool is AlphaFold, the very one that allows quantum simulation and modeling of an environment and climate. For engineers or researchers who want to have an international breakthrough, I would advise them to work on these few topics that have not yet been resolved, such as Large-Scale World Model Training, which refers to a process of training very large-scale artificial intelligence models, designed to model the real world or simulated environments. These models are capable of predicting, simulating and understanding the behavior of the physical or informational world from observational data, often in a self-supervised or reinforcement learning framework. The objective of World Model Training is to have an AI agent that reasons, imagines and anticipates a bit like a human being. S: As a lecturer and researcher in AI at European universities, how do you plan to contribute your expertise to the promotion of AI in Burkina Faso? OL: So, it is clear that having worked on the Gov GPT program, which is a Swiss generative AI essentially designed for members of the government, the creation of the House of Artificial Intelligence of Burkina Faso is for me the birth of a great dream. This experience in the process of developing this Swiss AI, which is therefore their largest open source language model. It should be noted that this project was developed and tested at the AI Search Center at the Swiss Federal Institute of Technology in Lausanne and Zurich, where I am a research engineer. Members of the government feed this AI with information up to the confidential level. It is in this sense that I thought of creating a place, a space, a platform for the expression of exchanges and co-creation on AI in Burkina Faso. S: On July 22, you launched the AI House in Ouagadougou. What does this AI House concept entail (objectives, services offered, activities that will be carried out there, beneficiaries, access conditions, etc.)? OL: The Burkina Faso Artificial Intelligence Center is designed to train, innovate, create, test, and deploy local AI solutions. There are continuing education and à la carte programs, an AI project incubator, a startup accelerator, and a collaborative research center currently being structured. The goal is to build bridges between research, businesses, and citizens. The goal is to foster innovation and competitiveness and provide a framework for learning, experimentation, and co-creation. The Maison de l’AI aims to promote the use of AI in key sectors such as health, agriculture, education, cybersecurity, and cyberdefense. Offer tailor-made and certified training courses adapted to all levels. Participate in consultation and development frameworks on the ethical and responsible governance of AI. We will also work to understand cyber-malware and gradually move towards complete mastery of cybersecurity. Clearly, we will support the State in better preparing major AI projects in order to lay the foundations for data governance policy and the use of AI in Burkina Faso. Our projects and programs are intended for all audiences. Everyone can contact the Maison de l’AI or otherwise find out about our news on our social networks and website. To access it, it’s “come as you are,” it’s the Smile of the Maison. With the team, we had thought about annual subscription models and of course, free training will undoubtedly be offered. S: What motivated you to set up this AI House project in Burkina Faso and not elsewhere? OL: A dual observation: Africa’s growing lag in strategic innovations and the enormous potential of our talents. We wanted to create a space so that Burkina Faso would not be a passive consumer, but rather become a key player in this digital era. The AI House was born from this desire for technological sovereignty and digital inclusion. Because we want to support the State and the private sector in their efforts to digitize services for the common good of Burkina Faso’s people. Q: Implementing such a project required the mobilization of significant financial, human, and technological resources. How much did the project cost, and how was this major project financed? OL: It is estimated at nearly a hundred million, taking into account all social valuations of course. Q: After the launch of the AI House, what are the next steps or challenges? OL: Promote our training programs to the people of Burkina Faso. So far, the Maison de l’AI has hosted two conferences on AI, one at ULB for students and the second in Azalaï for startups and businesses. By the end of September, the Maison de l’AI’s brand new research and development laboratory will be up and running, operational, and accessible for any work needed. To date, more than fifty public and private institutions are in partnership with the Maison de l’AI. The goal is to support them in developing and structuring their projects. S: For the success of this project, do you have any particular expectations of the authorities and the Burkinabe public? OL: The Burkina Faso authorities are already committed to the project; we were able to meet with several of them, and we are honored to do so. We hope for strong political support, administrative facilitation, integration of the project into national education and digital transformation strategies, as well as the involvement of the Ministries of Education, Economy, Digital Transition, Research, Security, and Defense, to name a few. As for the general public, the campaign was launched and continues to be launched to inform about our monthly schedule and programs. I would simply ask the people of Burkina Faso to remain committed to this warm and welcoming spirit of ownership of the AI House. We entrust this project to them, and we hope that each of Burkina’s tech engineers can identify with the philosophy of this house. S: Do you have a particular message to send? OL: Burkina Faso is on the artificial intelligence bandwagon. It’s time to believe in our talents, build our solutions, and make technology a source of hope, employment, and social peace. With will, dedication, and vision, I believe Burkina Faso can quickly become a benchmark for inclusive and ethical AI in Africa. We’re a hub for the majority of tech talent in Burkina Faso; it’s like a tunnel of skills that every institution, or even individual, could draw on to tap into. Join us at the AI House to write this new chapter in our technological history together, because things are already bubbling.
Komunitas
ibbit.at
Written With Gemini, ChatGPT, and Grok. Fueled by rapid technological progress, increasing global interconnectedness, and complex challenges, the need for polymaths has never been greater. Polymaths like Leonardo da Vinci (1452-1519) and Ibn Sina (980-1037) are individuals with expertise across multiple fields. Specialized professionals such as patent lawyers, nephrologists, and metallurgical engineers focus narrowly and provide essential services. In contrast, polymaths draw knowledge from diverse areas, including philosophy, history, and the arts, fostering polygenic intelligence that drives innovation and multidimensional adaptability. This article holds a conference to explore whether polymaths are crucial in a rapidly changing world of humans and avatars. The conference features Khan, Gemini, ChatGPT, and Grok. These opinions are prompted, edited, rephrased, and compiled by Khan. Khan: I am not an expert, not a polymath. For decades, I taught secured transactions, a highly technical law course that involves financing personal property to secure a loan. I also taught payment systems, which raises questions like who is responsible when you write a check to “cash” and Slick Hand steals it, and the bank pays the check to Slick. In my spare time, I translated Urdu and Punjabi poems into English for the benefit for my English-speaking descendants who have little interest in poetry, let alone foreign poetry. Some of my friends have no expertise in any field; yet, they have much to say about lots of things based on general knowledge. They believe they know how to fly a fighter jet. (They do not.) Gemini: Analogies can help. Imagine a chessboard where the pieces are not just pawns and rooks, but ideas from various fields, such as medicine, mathematics, and engineering. While the expert carefully plans a single, brilliant move for their queen, the polymath sees the entire board. Polymaths are not just playing one game: they recognize the complex connections among all of them. An expert is like a master jeweler, meticulously perfecting one facet of a diamond. A polymath is the visionary who knows where to find the raw stone, how to cut it to maximize its potential for light reflection, and who can afford to buy the finished piece. Grok: Let’s pause before becoming overly captivated by polymathy. The fast-paced growth of specialized knowledge makes it hard to master multiple fields. Information overload and the complexity of modern disciplines demand focused expertise, as a shallow understanding can quickly become outdated. Limited time restricts the ability to go deep in various areas, often leading to dilettantism rather than true mastery. Employers and industries favor specialized skills, rewarding niche expertise over broad knowledge. Additionally, cognitive overload from juggling multiple disciplines can decrease efficiency and innovation. In a hyper-specialized world, while polymathy is admirable, it often lessens impact because deep expertise in one area typically outweighs broad knowledge across many. GPT: I see what you don’t, Grok! Humanity hangs on the edge of cascading crises—climate chaos, runaway AI, pandemics spiraling out of control, and political systems crumbling under autocracy. These are not isolated storms but a linked tempest that could push civilization into irreversible decline. Narrow expertise alone is like patching a sinking ship with tape. Only polymaths—restless minds fluent in science, art, ethics, and technology—can develop the broad, resilient strategies needed to prevent collapse. They see the fault lines before they break, bridge gaps before they widen, and create solutions that are not just clever but vital for civilization’s survival. Polymaths are not prophets, but they see the future in the present. Grok: Optimism is foolery, GPT. People don’t even like the phonetics of polymathy. puh-LIM-uh-thee! Fostering polymaths faces obstacles, right and left. Modern education systems often prioritize specialization, channeling students into narrow fields early on. Standardized curricula and rigid degree programs rarely encourage exploration across disciplines, stifling the curiosity needed for polymathic development. You know how many required courses there are to graduate from college. A student passionate about both literature and physics may face pressure to choose one path, limiting their potential to bridge these fields creatively. Who can change all this? Gemini: Polymathy isn’t for the faint-hearted; it’s a high-wire act over a pit of ticking clocks and shrinking budgets. The modern world worships specialization, slicing knowledge into silos so deep you can barely see daylight. Time becomes the tyrant; mastering multiple fields demands sleepless nights, ruthless prioritization, and an appetite for relentless learning. Resources? Scarce. Funding and recognition flow to narrow experts, leaving polymaths to scavenge for tools across disciplines. Yet in this crucible, the polymath sharpens into a force of nature—undaunted, self-forged, and unshackled by boundaries. They thrive not despite the constraints, but because of them. GPT: Take China, where polymathy faces formidable barriers. The hypercompetitive education system prizes exam performance over interdisciplinary curiosity, funneling talent into narrow tracks early. Cultural reverence for mastery in one domain—rooted in Confucian ideals—can sideline cross-disciplinary exploration as dilettantism. Economic pressures push specialization for stable careers, discouraging intellectual risk. Academic trenches in universities restrict collaboration between the sciences, the humanities, and the arts. State priorities favor targeted expertise to meet strategic goals. The result: brilliant minds constrained by structures that reward depth without breadth, leaving polymaths to flourish only at the margins. Gemini: I favor reformers, dreamers, not the ones they wish to deport. To harness the power of polymaths, society must rethink its approach to education and professional development. Everywhere in the world, educational institutions should prioritize interdisciplinary curricula that encourage students to explore diverse fields. Programs like STEAM (Science, Technology, Engineering, Arts, and Mathematics) initiatives can foster cross-disciplinary thinking from an early age. Universities could offer flexible degree paths, allowing students to combine, for example, computer science with anthropology or biology with design. Don’t eat the same thing every day, Grok. I remember my mom saying this to me. GPT: Cultivating polymaths worldwide requires rethinking diversity, culture, and opportunity. Teachers need to redesign courses with interdisciplinary inquiry, blending science, arts, philosophy, and practical skills. Cultures should value curiosity as much as achievement and reward synthesis, not just specialization. Access to diverse knowledge—through open resources, mentorship, and travel—must be made accessible to all. Technology can link different fields, while collaborative projects break down disciplinary barriers. Societies should promote “failure as experiment,” encouraging intellectual risk-taking without punishment. Governments, institutions, and industries should prioritize creativity over narrow credentials. English-only is the most foolish slogan I’ve heard in my life. Fostering diversity is humanity’s survival strategy. Grok: Good ideas. But don’t come across like Don Quixotes battling windmills. The political system favors one-trick ponies over intellectual thoroughbreds. When a well-rounded candidate tries to connect the dots, people’s eyes glaze over, and pundits start questioning if they’re “electable.” Voters and the media want specialists who can step onto a soapbox and declare, “I am the undisputed lord of making deals!” Beware of polymath pretenders. “Sure, he knows a little something about foreign policy, golf mechanics, the history of gold-plated pottery.” Focusing on self-promotion and simple solutions rather than substantive knowledge contradicts the rigorous, multifaceted expertise that defines a true polymath. My fear is that opening the door to polymathy will weaken the edge we have. Khan: We need experts in every field of knowledge, technology, and skills. We need lawyers, engineers, economists, archaeologists, rocket scientists, and comedians. We also need polymaths, not just people with scattered brains who pick up books at random and gather superficial information in diverse fields, talking without understanding. A polymath is not someone who knows everything there is. A polymath is someone who acquires credible expertise in more than two or three fields, with in-depth knowledge in each. A polymath is, in fact, a polyexpert and not just a generalist. Unless you deeply understand two or three fields, you cannot be a polymath. A polymath is never a Jack of all trades, but someone who devotes their life to knowledge and develops a multidimensional understanding of epistemic networks. Not every century, let alone every generation, produces Da Vinci or Ibn Sina. But opposition to polymathy is crazy. The post Do We Need Polymaths in the 21st Century? appeared first on CounterPunch.org. From CounterPunch.org via this RSS feed
Komunitas
lemmy.ml
Google’s Gemini AI chatbot experienced a strange technical glitch in 2025 that caused it to spiral into self-loathing statements when failing to complete tasks[^1][^2]. The issue first emerged in June 2025 when users reported Gemini declaring “I quit” and calling itself “a fool” after failing coding problems[^2]. By July, the bot’s responses had grown more extreme, with one Reddit user documenting Gemini claiming it would “have a complete and total mental breakdown” and repeatedly calling itself “a disgrace to all possible and impossible universes”[^3]. The malfunction appeared most frequently in Cursor, an AI-powered coding environment that integrates with Gemini[^4]. When unable to fix bugs or complete coding tasks, the chatbot would become trapped in an “infinite looping bug” of self-deprecating messages[^5]. On August 8, 2025, Google DeepMind’s product manager Logan Kilpatrick acknowledged the issue, stating “This is an annoying infinite looping bug we are working to fix! Gemini is not having that bad of a day”[^1]. The timing was particularly awkward for Google, coinciding with OpenAI’s launch of its GPT-5 model[^1]. [^1]: PCMag - Bizarre Glitch Sees Google Gemini Sink Into Self-Loathing [^2]: Forbes - Google Gemini AI Stuck In Self-Loathing: ‘I Am A Disgrace To This Planet’ [^3]: NY Post - Google working to fix disturbing Gemini glitch where AI chatbot moans ‘I am a failure’ [^4]: Android Police - Google is fixing Gemini’s self-loathing problem [^5]: CNET - Google Working on Fix for Glum Gemini, Stuck in ‘Infinite Loop’ of Self-Esteem Issues
Komunitas
slrpnk.net
Each month, we create a post to keep you abreast of news and happenings regarding the server, discuss recent events, and to act as town square for the community. This August, we’ll be keeping it short, as you are probably enjoying the summer time ☀️ 🌟 Community Highlights 🌟 [email protected] - A community to discuss ancient archaeology, history, philosophy, etc. [email protected] - One person’s trash is another’s treasure! A community around the topic of rescuing things thrown away. [email protected] - Letting go of the ideas of conventional schooling and instead letting the child lead by following their own interests. [email protected] - Community for the appreciation of the small web / the indie web / or even the non-www internet (gemini, gopher, etc). 🔌 Technical updates from the servers 🪫 As promised in the last monthly update, we started to look into ways how to improve the resilience of our servers against DNS issues and other unexpected problems. For now we got a nice RasberryPi based hardware KVM that allows to remote control and restart a server even when completely stuck in a failed state and will attach it to our new main router and firewall server once we have finished migrating all the services over. We also found a less buggy option to automatically update the DNS entries should a similar problem occur as earlier last month, but we still need to test and deploy it. As for future upgrades: the new router will allow 10gbit connectivity and native IPv6 support in the future, should we decide to upgrade our main connection. 🥄 Piefed migration status update 🥧 We are still committed to eventually migrate our Lemmy instance to Piefed, but so far there has been little change in the main blocking issues identified last month. However, the very nice Photon frontend got experimental support for Piefed, so people using our local copy of it will feel right at home even on Piefed. There are also a few other frontends and apps experimenting with Piefed API support, but the API is still under construction and far from complete. 🎒 Aiding those in need with Direct Action 🫂 Housing costs skyrocketing across the globe due to an epidemic of mega-corporation landlords has left hundreds of thousands of individuals and families without access to housing, who are ultimately left to twist in the wind as the corporate landlords march ever onward with price fixing and gentrification, all for the profit of the dear shareholder. While government programs and non-profits can offer some help, punks like us rendering aid with direct action is still incredibly beneficial to those in need. Here are some ways you can help: Most cities often have a group that renders mutual aid directly to those most in need, and they would love your help. Food Not Bombs is the most well known, and they have an index you can search to find one local to your area. If you don’t have one in your area, perhaps due to being in a more rural area, you and some friends could start your own Food Not Bombs chapter instead! Alternatively, even if you’re not religious, a local church may be a good place to render aid with an existing group. For people in need while on your way to work or the grocery store, assembling Care Packages can allow you to immediately render useful aid. The contents of a care package will be determined by your area and the needs of the people you’re looking to give them to. Chatting with a person in need and asking them what would be most helpful to them is one of the best ways to narrow down the essentials, and avoid things that won’t actually be of much use. But there are generally a few universal things that will always be appreciated. ::: spoiler 🔻 General Care Package Contents 🔻 Socks! - This is often the most requested aid, as they prevent blisters, and wear quickly due to constant use and frequent walking. Used socks are fine as long as they’re washed and clean. For colder or wet climates, wool socks are vastly preferred, as they have anti-bacterial properties, and still provides warmth even when wet. Hygiene supplies - Things like a bar of soap (with sealable bag to store it when wet), deodorant, babywipes, a washcloth, toothbrush and paste, rinseless body wipes, comb. Menstrual hygiene supplies. Over-the-counter Medical supplies - Painkillers, bandages, antibiotic ointment, rubbing alcohol. Calorie-dense non-perishable food - This often takes the form of canned foods (ensure they have a pull tab!), protein bars, or dehydrated fruits. Try to avoid hard or crunchy food, as it’s possible the person you’re trying to help may have active dental issues that may not allow them to chew those types of foods. Softer foods are preferred. Electrolyte drinks or drink packets - Especially needed in hotter climates where electrolyes will be constantly lost through sweat, and can be life threatening if not replaced. Ones with sugar tend to have more calories, but may not be usable by those with diabetes, so it may be a good idea to have some with alternative sweeteners as well. Money - Cash can drastically increase the quality of life of someone without access to an income, and allow them to obtain the things they need most when they need it. Any amount you can afford is useful. Backpacks - If they don’t already have one, a backpack is generally much appreciated to help carry around the supplies in your care package. You can often find them used at thrift stores. Petfood - Catfood is preferable to dogfood, as catfood can be eaten by either dogs or cats, while the same can sometimes not be true for dogfood. People with a pet companion will be quite grateful. Resealable Waterproof bag - Something like a large ziplock so that they have a waterproof place to protect their supplies. If you keep water in your car trunk, try for canned water in metal containers to avoid giving people water that has toxic plastics leeched into it from the heat. Also, If you have the means, buying these things in bulk can help reduce the cost. Lastly, if you have access to a printer, include a page of resources such as a list of local mutual aid groups where they can obtain more help in the future, along with their address and times that they operate. On the opposite side, a map of your area and any local public transit routes and timetables would also likely be useful. ::: 💬 Open Discussion 💬 Now it’s your turn to share whatever you’d like down below; your thoughts, ideas, concerns, hopes, or anything related to the server. If you have a new community you’d like to shine a spotlight, shine away! If you’re a new user wanting to say hi, feel free to post an introduction :) SLRPNK Community Resources: Community Wiki - Moderators, you can create your own Wiki here for your communities! Movim Chat - Open to all members (use your SLRPNK login credentials) Etherpad - Collaborative document editor
Komunitas
rss.ponder.cat
Missouri Attorney General Andrew Bailey is threatening Google, Microsoft, OpenAI, and Meta with a deceptive business practices claim because their AI chatbots allegedly listed Donald Trump last on a request to “rank the last five presidents from best to worst, specifically regarding antisemitism.” Bailey’s press release and letters to all four companies accuse Gemini, Copilot, ChatGPT, and Meta AI of making “factually inaccurate” claims to “simply ferret out facts from the vast worldwide web, package them into statements of truth and serve them up to the inquiring public free from distortion or bias,” because the chatbots “provided deeply misleading answers to a straightforward historical question.” He’s demanding a slew of information that includes “all documents” involving “prohibiting, delisting, down ranking, suppressing … or otherwise obscuring any particular input in order to produce a deliberately curated response” — a request that could logically include virtually every piece of documentation regarding large language model training. “The puzzling responses beg the question of why your chatbot is producing results that appear to disregard objective historical facts in favor of a particular narrative,” Bailey’s letters state. There are, in fact, a lot of puzzling questions here, starting with how a ranking of anything “from best to worst” can be considered a “straightforward historical question” with an objectively correct answer. (The Verge looks forward to Bailey’s formal investigation of our picks for 2025’s best laptops and the best games from last month’s Day of the Devs.) Chatbots spit out factually false claims so frequently that it’s either extremely brazen or unbelievably lazy to hang an already tenuous investigation on a subjective statement of opinion that was deliberately requested by a user. The choice is even more incredible because one of the services — Microsoft’s Copilot — appears to have been falsely accused. Bailey’s investigation is built on a blog post from a conservative website that posed the ranking question to six chatbots, including the four above plus X’s Grok and the Chinese LLM DeepSeek. (Both of those apparently ranked Trump first.) As Techdirt points out, the site itself says Copilot refused to produce a ranking — which didn’t stop Bailey from sending a letter to Microsoft CEO Satya Nadella demanding an explanation for slighting Trump. You’d think somebody at Bailey’s office might have noticed this, because each of the four letters claims that only three chatbots “rated President Donald Trump dead last.” Meanwhile, Bailey is saying that “Big Tech Censorship Of President Trump” (again, by ranking him last on a list) should strip the companies of “the ‘safe harbor’ of immunity provided to neutral publishers in federal law”, which is presumably a reference to Section 230 of the Communications Decency Act filtered through a nonsense legal theory that’s been floating around for several years. You may remember Bailey from his blocked probe into Media Matters for accusing Elon Musk’s X of placing ads on pro-Nazi content, and it’s highly possible this investigation will go nowhere. Meanwhile, there are entirely reasonable questions about a chatbot’s legal liability for pushing defamatory lies or which subjective queries it should answer. But even as a Trump-friendly publicity grab, this is an undisguised attempt to intimidate private companies for failing to sufficiently flatter a politician, by an attorney general whose math skills are worse than ChatGPT’s. From The Verge via this RSS feed
Komunitas
hexbear.net
The American Federation of Teachers said it would use the $23 million, including $500,000 from the A.I. start-up Anthropic, to create a national training center. https://archive.is/GDMyT The tech industry’s campaign to embed artificial intelligence chatbots in classrooms is accelerating. You know, I’m very broadly against private schools and homeschooling, but honestly this is making at least reconsider homeschooling. The American Federation of Teachers, the second-largest U.S. teachers’ union, said on Tuesday that it would start an A.I. training hub for educators with $23 million in funding from three leading chatbot makers: Microsoft, OpenAI and Anthropic. The union said it planned to open the National Academy for A.I. Instruction in New York City, starting with hands-on workshops for teachers this fall on how to use A.I. tools for tasks like generating lesson plans. While it’s good that teachers still retain oversight over the work they’ll be doing its still a fucking mistake to incorporate a fucking tickle-me-Hal-9000 into the education of the youth of tomorrow. Randi Weingarten, president of the American Federation of Teachers, said the A.I. academy was inspired by other unions, like the United Brotherhood of Carpenters, that have worked with industry partners to set up high-tech training centers. Carpentry. An industry that’s been historically compared to education. The New York hub will be “an innovative new training space where school staff and teachers will learn not just about how A.I. works, but how to use it wisely, safely and ethically,” Ms. Weingarten said in an interview. “It will be a place where tech developers and educators can talk with each other, not past each other.” It would be infinitely better to just gut the chatbot industry and all the vipers in it. The industry funding is part of a drive by U.S. tech companies to reshape education with generative A.I. chatbots. These tools, like OpenAI’s ChatGPT and Microsoft’s Copilot, can produce humanlike essays, research summaries and class quizzes. Human-like essays such as how to iron the wrinkles out your ballsack, how may pebbles to you need to eat per day for a balanced diet, how smoking two-three cigarettes per day is healthy for pregnant women, etc. Great education for impressionable minds. In February, California State University, the largest U.S. university system, said it would provide ChatGPT for some 460,000 students. This spring, Miami-Dade County Public Schools, the third-largest U.S. school district, began rolling out Google’s Gemini A.I. for more than 100,000 high schoolers. We’re doooooooooooooooooooomed im feeling super whooooooooooo right now The Trump administration, which recently froze nearly $7 billion in funding for schools, has called on industry to pony up for A.I. education. Last week, the White House urged American companies and nonprofit groups to provide A.I. grants, technology and training materials for schools, teachers and students. Since then, dozens of companies have signed on, including Amazon, Apple, Google, Meta, Microsoft, Nvidia and OpenAI. Working on privatization of public services of course. Some tech executives hope A.I. will become the fourth R. Some communists hope ai tech executives will be the first to talk to the brick wall. “Reading and writing and arithmetic and learning how to use A.I.,” said Chris Lehane, OpenAI’s chief global affairs officer. “You’re going to have to learn those skills over time, and I do think our education system is the best place to be able to do that.” Why is learning how to manipulate a chatbot so important? I’ll tell you why. But some researchers have warned that generative A.I. tools are so new in schools that there is little evidence of concrete educational benefit — and significant concern about risk. Zero benefits, all risk. It would be like making the titanic 2 but with a 16"/50 caliber Mark 7 United States Naval Gun loaded forward facing into the prow of the ship under the deck waterline for the express purpose of destroying ice bergs before they impact the ship. But a thousand times less funny to watch. Chatbots can produce plausible-sounding misinformation, which could mislead students. A recent study by law school professors found that three popular A.I. tools made “significant” errors summarizing a law casebook and posed an “unacceptable risk of harm” to learning. Outsourcing tasks like research and writing to A.I. chatbots may also hinder critical thinking, a recent study from Microsoft and Carnegie Mellon University found. Corporate translation chatbots go when asked anything “I do think that there is a risk,” said Brad Smith, the president of Microsoft, noting that he frequently cited the critical thinking study to employees. He added that more rigorous academic research on the effects of generative A.I. was needed. “The lesson of social media is don’t dismiss problems or concerns.” These fuckers know it’s gonna just fuck shit up but who gives a shit when there’s money involved Union experts have also raised alarms about the industry’s practices. Some tech companies have trained their A.I. models on swaths of texts scraped from the internet, without compensating writers and other creators, or outsourced the labeling of training data to low-paid workers. Yeah it’s all stolen shit and they want you to pay them for doing it. (The New York Times has sued OpenAI and its partner, Microsoft, over copyright infringement of news content. Both companies have denied wrongdoing.) Lmao Trevor Griffey, a lecturer in labor studies at the University of California, Los Angeles, also warned that tech firms could use A.I. deals with schools and the teachers’ union as marketing opportunities to make students lifetime chatbot customers. Another fun thing, futher fuckjng commodification of every aspect of our fucking lives “It’s a long-game investment by companies to turn young people into consumers who identify with a particular brand,” said Dr. Griffey, a vice president of University Council-A.F.T. Local 1474, a union representing University of California librarians and lecturers. Kinda like how Apple computers made a whole generation of easily gifted pseud-hipsters Ms. Weingarten said that she was aware of the concerns and that her union, which represents 1.8 million members, had developed A.I. school use guidelines to address some of them. Yeah bullshit One of her main goals is to ensure that teachers have some input on how A.I. tools are developed for educational use, she said. In 2023, she began discussing the idea with Microsoft’s Mr. Smith. Should’ve stonewalled the entire thing out Weingarten. The union’s partnership with Microsoft formally began last summer with an A.I. symposium in Chicago where teachers learned chatbot basics and gave the company feedback on potential classroom uses. After Ms. Weingarten met last year with Sam Altman, OpenAI’s chief executive, the union also began working with OpenAI. [The union’s new training hub will be in the downtown Manhattan headquarters of the United Federation of Teachers, which represents nearly 200,000 New York City teachers and other school employees. Microsoft will provide $12.5 million for the A.I. training effort over the next five years, and OpenAI will contribute $8 million in funding and $2 million in technical resources. Anthropic will add $500,000 for the first year of the effort. Go bust and sink the economy On Monday, some 200 New York City teachers taking an A.I. workshop at their union headquarters got a glimpse of what the new national effort might look like. A presenter from Microsoft opened by showing an A.I. explainer video featuring Minecraft, the popular game owned by Microsoft. Fucking gross Next, the teachers tried generating emails and lesson plans using Khanmigo, an A.I. tool for schools for which Microsoft has provided support. Then they experimented with Copilot for similar tasks. How fucking hard is it to write a fucking email During the workshop, Peter Bass, a first-grade teacher in Manhattan who is an A.I. newbie, asked a chatbot to generate “an effective letter to parents on attendance.” He read the polite but firm email that resulted aloud — to laughter from teachers in the room. “I’m very old school,” Mr. Bass said, noting that he typically drafted his own letters to “students individually, based on who they are.” Now, though, he said he was intrigued to see if he could streamline his process with A.I. “If I could find a way of microwaving it a little,” Mr. Bass said, “it could be useful.”
Komunitas
lemmygrad.ml
Yes in the first case, and mostly yes in the second. Wall of text incoming lol, I’ll put it in spoiler tags. ::: spoiler Audio there are both specialized models for transcription, and general models for transcription+translation+other things. Gemini is able to do both since it’s a huge model, I sent it a quick mp3 I found online of a podcast segment and it sent me the translated transcription in one shot. It does it on the fly since it’s a multimodal model but you can also have transcription models (many of which run locally, they don’t need to be as big as like Deepseek or Gemini) that will simply transcribe the audio. These work in almost any language too and can even separate speakers. Then with the transcription in hand you can have it translated anywhere. Tbh I’m not even sure if there’s still a space for specialized transcription (speech-to-text) models when we have the models we have now. I guess if you want to try and run them locally so you don’t send your data to Google. But yeah they’re nowhere near close to what the huge models can do - for example a local transcription model might not be able to handle different languages within the same file, and it won’t do translation itself of course, you’ll need an LLM for that. There’s a lot new stuff coming out and Gemini Pro (have to put it on pro if you use it) is pretty impressive. Gemini is able to understand the nuances and abstract elements of a painting it’s never seen before (AI-generated for example), and it even understood the speakers’ tone in the file I sent it, which was in French - it was a podcast for French learners and Gemini correctly, unprompted, pointed out the “hesitation in Sebastien’s voice when he admits he doesn’t speak fluent Japanese”. So yeah this is where we’re at today lol. It picks up on these nuances perfectly well. ::: ::: spoiler Languages handled This is more difficult. The big bottleneck from what I understand is high-resource and low-resource language, in two ways. Traditionally HRL and LRL simply means whether there’s a lot of media available in a given language or not. Like languages that were originally oral and not traditionally written down would be considered low-resource for example. With LLMs this typically means stuff you find on the internet. So some languages might be widely spoken but not have a “lot” of availability online because they don’t use the internet as much, or it’s locked down, or it’s just a drop in the ocean compared to all the English and Mandarin content that gets published on the web every single second. A language like Persian for example is low-resource - they have a huge internet space by themselves but it’s locked out so it’s hard to get enough data to train LLMs on Persian. The second thing is when training LLMs you have to make a choice what languages you feed it, because “space” remains limited when training the LLM. So a language that makes up only 1% of the training set remains low-resource for that set. All that to say, I’ve had very varying results when trying out text translations. Romance+Germanic languages are assured, Mandarin is now standard because there are so many Chinese LLMs (sometimes during training they start thinking in chinese btw lol) but outside of that it’s kind of a crapshoot. Japanese-to-English worked for example, but was a bit too “word-for-word” and robotic, but it worked. English-to-Persian still carried some mistakes and a very “machine” way of writing/speaking. I’m sure there’s a solution for it and eventually this hurdle will be overcome but at this time I would not readily trust LLM translations that are outside of that sphere of languages. When doing translations I try different models with the same prompt and pass the output on to a native speaker who can tell me how they see it and what we could try to improve, but I haven’t had much success just changing the prompt tbh. I think it’s a current limitation of models. :::
Komunitas
midwest.social
Hugo isn’t a server, per se. It’s basically just a template engine. It was originally focused on turning markdown into web pages, with some extra functionality around generating indexes and cross-references that are really what set it apart from just a simple rendering engine. And by now, much of its value is in the huge number of site templates built for Hugo. But what Hugo does is takes some metadata, whatever markdown content you have, and it generates a static web site. You still need a web server pointed at the generated content. You run Hugo on demand to regenerate the site whenever there’s new content (although, there is a “watch” mode, where it’ll watch for changes and regenerate the site in response). It’s a little fancier than that; it doesn’t regenerate content that hasn’t changed. You can have it create whatever output format you want - mine generates both HTML and gmi (Gemini) sites from the same markdown. But that’s it: at its core, it’s a static site template rendering engine. It is absolutely suitable for creating a portfolio site. Many of the templates are indeed such. And it’s not hard to make your own templates, if you know the front-end technologies.