Latest news with #PaulGraham
Business Insider
a day ago
- Business
- Business Insider
Thanks to ChatGPT, the pure internet is gone. Did anyone save a copy?
In the post-nuclear age, scientists noticed a peculiar problem: steel produced after 1945 was contaminated. Atomic bombs had infused the atmosphere with radioactivity, which contaminated the metal. This made most steel useless for precise equipment such as Geiger counters and other highly accurate sensors. The solution? Salvage old steel from sunken pre-war battleships resting deep on the ocean floor, far away from the nuclear fallout. This material, known as low-background steel, became prized for its purity and rarity. Fast forward to 2025, and a similar story is unfolding — not under the sea, but across the internet. Since the launch of ChatGPT in late 2022, AI-generated content has exploded across blogs, search engines, and social media. The digital realm is increasingly infused with content not written by humans, but synthesized by models and chatbots. And just like radiation, this content is tricky for regular folks to detect, is pervasive, and it alters the environment in which it exists. This phenomenon poses a particularly thorny problem for AI researchers and developers. Most AI models are trained on vast datasets collected from the web. Historically, that meant learning from human data: messy, insightful, biased, poetic, and occasionally brilliant. But if today's AI is trained on yesterday's AI-generated text, which was itself trained on last week's AI content, then models risk folding in on themselves, diluting originality and nuance in what's been dubbed " model collapse." Put another way: AI models are supposed to be trained to understand how humans think. If they're trained mostly on their own outputs, they may end up just mimicking themselves. Like photocopying a photocopy, each generation becomes a little blurrier until nuance, outliers, and genuine novelty disappear. This makes human-generated content, from before 2022, more valuable because it grounds AI models, and society in general, in a shared reality, according to Will Allen, a vice president at Cloudflare, which operates one of the largest networks on the internet. This becomes especially important as AI models spread into technical fields, such as medicine, law, and tax. He wants his doctor to rely on content based on research written by human experts from real human trials, not AI-generated sources, for instance. "The data that has that connection to reality has always been critically important and will be even more crucial in the future," Allen said. "If you don't have that foundational truth, it just becomes so much more complicated." Paul Graham's problem This isn't just theoretical. Problems are already cropping up in the real world. Almost a year after ChatGPT launched, venture capitalist Paul Graham described searching online for how hot to set a pizza oven. He found himself looking at the dates of the content to find older information that wasn't " AI-generated SEO-bait," he said in a post on X. Malte Ubl, CTO of AI startup Vercel and a former Google Search engineer, replied, saying Graham was filtering the internet for content that was "pre-AI-contamination." "The analogy I've been using is low background steel, which was made before the first nuclear tests," Ubl said. Matt Rickard, another former Google engineer, concurred. In a blog post from June 2023, he wrote that modern datasets are getting contaminated. "AI models are trained on the internet. More and more of that content is being generated by AI models," Rickard explained. "Output from AI models is relatively undetectable. Finding training data unmodified by AI will be tougher and tougher." The digital version of low-background steel The answer, some argue, lies in preserving digital versions of low-background steel: human-generated data from before the AI boom. Think of it as the internet's digital bedrock, created not by machines but by people with intent and context. One such preservationist is John Graham-Cumming, a Cloudflare board member and the company's CTO. His project, catalogs datasets, websites, and media that existed before 2022, the year ChatGPT sparked the generative AI content explosion. For instance, there's GitHub's Arctic Code Vault, an archive of open-source software buried in a decommissioned coal mine in Norway. It was captured in February 2020, about a year before the AI-assisted coding boom got going. Graham-Cumming's initiative is an effort to archive content that reflects the web in its raw, human-authored form, uncontaminated by LLM-generated filler and SEO-optimized sludge. Another source he lists is "wordfreq," a project to track the frequency of words used online. Linguist Robyn Speer maintained this, but stopped in 2021. "Generative AI has polluted the data," she wrote in a 2024 update on coding platform GitHub. This skews internet data to make it a less reliable guide to how humans write and think. Speer cited one example that showed how ChatGPT is obsessed with the word "delve" in a way that people never have been. This has caused the word to appear way more often online in recent years. (A more recent example is ChatGPT's love of the em dash — don't ask me why!) Our shared reality As Cloudflare's Allen explained, AI models trained partly on synthetic content can accelerate productivity and remove tedium from creative work and other tasks. He's a fan and regular user of ChatGPT, Google's Gemini, and other chatbots such as Claude. And just like human-generated data, the analogy to low-background steel is not perfect. Scientists have developed different ways to produce steel that use pure oxygen. Still, Allen says, "you always want to be grounded in some level of truth." The stakes go beyond model performance. They reach into the fabric of our shared reality. Just as scientists trusted low-background steel for precise measurements, we may come to rely on carefully preserved pre-AI content to gauge the true state of the human mind — to understand how we think, reason, and communicate before the age of machines that mimic us. The pure internet is gone. Thankfully, some people are saving copies. And like the divers salvaging steel from the ocean floor, they remind us: Preserving the past may be the only way to build a trustworthy future.
Yahoo
4 days ago
- Entertainment
- Yahoo
Musk Blows Up Online as ‘Ketamine' Dinner Clip Resurfaces
Elon Musk issued a curt rebuke of ketamine abuse allegations after a viral clip of his erratic behavior at a dinner recirculated online. The clip—taken in March during a visit to President Donald Trump's Bedminster golf club in New Jersey—shows Musk balancing spoons at a dinner table while one of his baby mamas, Shivon Zilis, watches in silent concern. At the time, X user and 'retired army medic' Molly Ploofkins tweeted the clip with the caption, 'Musk playing with his silverware while tripping on ketamine at Bedminster.' Amid an explosive New York Times report accusing Musk of using drugs as his Department of Government Efficiency bulldozed its way across the federal government, computer scientist and writer Paul Graham attempted to normalize the clip with a tweet insisting that he does the same with his silverware. Responding to the original clip, Graham tweeted, 'You don't need ketamine to make these. I always make them. In our family we call them 'Robert towers,' after Robert Morris, who also makes them.' The remark elicited a visceral response from Musk: 'I'm not on ketamine ffs,' he wrote. Graham responded under his comment, 'That's what I'm saying. Making cutlery towers isn't evidence that someone's on drugs. You could take an identical video of me at a dinner.' As Trump bid adieu to Musk in a Friday press conference—in which the president insisted 'Elon is not really leaving'—the Tesla CEO shut down questions about the Times report. 'Is The New York Times the same publication that got a Pulitzer Prize for false reporting on Russiagate—is it the same organization?' Musk said. 'I think the judge just ruling against The New York Times for their lies about the Russiagate hoax and that they might have to give back that Pulitzer Prize. That New York Times? Let's move on!' Yet Musk's fierce denials contradict what he has publicly said and tweeted about using ketamine to counter his 'brain chemistry.' In a tweet posted Aug. 4, 2023, Musk responded to Nigerian rapper Zuby's concerns about antidepressants with his own take. 'I have serious concerns about SSRIs, as they tend to zombify people,' he wrote. 'Occasional use of Ketamine is a much better option, in my opinion. I have a prescription for when my brain chemistry sometimes goes super negative.' 'SSRIs' refers to 'selective serotonin reuptake inhibitors,' the class of antidepressants that includes Prozac and Zoloft. Musk echoed such sentiments again a year later in an interview with journalist Don Lemon, admitting to the former CNN host that he takes a 'small amount' of ketamine to treat depression. 'There are times when I have a sort of ... negative chemical state in my brain,' he explained. 'Ketamine is helpful for getting one out of a negative frame of mind.'
Yahoo
16-05-2025
- Business
- Yahoo
Google hit with scathing accusations from top VC firm
Google Parent Alphabet () faced a day of reckoning in April when the U.S. Department of Justice (DOJ) prevailed in a landmark case against it. So far, May 2025 isn't going any better. The tech-sector giant is still recovering from the recent ruling, in which the U.S. District Court for the Eastern District of Virginia ruled that its methods of monopolizing digital advertising markets had continuously violated antitrust law. 💵💰💰💵 This landmark decision shocked most of the tech world, which wondered if it would have further implications for the industry. Weeks later, the accusations against Google aren't stopping. An extremely prominent name in the venture capital (VC) space recently filed an amicus brief for the monopoly case, in which it levied harsh accusations against the Silicon Valley behemoth and speculated that it sees bigger problems ahead unless action is taken. Anyone who has founded a tech startup or simply follows the industry closely is likely familiar with the Y Combinator. A startup accelerator and venture capital firm, it has helped seed many of the sector's most promising new companies, including Airbnb, Instacart, and in 2005 by tech industry leaders such as Paul Graham, Jessica Livingstone, and Trevor Blackwell, Y Combinator has founded more than 5,000 startups. It's safe to say the company knows about innovation, which is why it finds Google's actions deeply concerning. In its amicus brief filed on May 9, a legal document filed by a third party with no direct involvement in the court case, Y Combinator laid out the problems it sees with Google's operations. One section highlighted the importance of open competitiveness in a free market and its consequences for a healthy, innovative ecosystem. As the startup accelerator's leaders see it, Google's way of conducting business has created a deadened area around innovation, described as the 'kill zone.' Because of this, VC firms have become reluctant to fund startups in the artificial intelligence (AI) and digital search spaces, due to the perceived risk of Google strangling new and early-stage companies in these fields. 'By Foreclosing competition, Google has chilled independent firms like YC from funding and accelerating innovative startups that could otherwise have challenged Google's dominance,' it states. 'The result is a landscape that has been artificially stunted and stagnant. In our view, Plaintiffs' proposed remedy package would help to unlock a more dynamic, globally competitive U.S. technology startup ecosystem.' More Google News: Apple exec shares a shocking take on Google's future Google sends a harsh message to employees after layoffs AI effect prompts Google parent stock-rating tweak Y Combinator also notes that in the current age of innovation, simply developing a new product is no longer enough. It argues that startup founders must be able to deliver what their company produces to its target user base, free from what it describes as the 'restrictive dealing and self-preferencing' created by Google. It seems that from the perspective of Y Combinator's leaders, judicial action against Google is necessary, as it would prevent Google from continuing to strangle innovation and negatively impact startup comes at an especially critical time, as more and more entrepreneurs look to break into the fast-growing areas such as AI, machine learning, and other areas that could fall within the parameters of the 'kill zone' noted by Y Combinator. However, Y Combinator CEO Garry Tan also noted that the company isn't advocating for breaking up Google. "We are not calling for a breakup of Google. If anything, we hope that a measured and thoughtful remedy between USG and Google from here (5-year term!) means a court-supervised spin-off is contingent, and a last resort — same playbook fixed AT&T and post-Microsoft era innovation," he recently stated in an X thread. It is clear, though, that the startup accelerator believes a more level playing field is essential for new companies entering the tech field and that Google's alleged monopolistic hold could severely compromise this, making things more difficult for both founders and venture capitalists who might be cautious about backing them. As the brief states, 'YC believes that this is a moment when dynamic competition could break out, but effective antitrust remedies in this case will be critical for the innovative startups currently trying to attract venture capital, reach users, and gain scale.'Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data
BBC News
01-04-2025
- Sport
- BBC News
'I study esports at school - it's not just playing games'
At well over six feet tall, you might think that Alfie would be more at home in the forwards pack on the rugby pitch when it comes to the 16-year-old Selkirk High School student is standing next to a bank of computer screens in front of hundreds of screaming fellow pupils crammed into the lunch part of a small team running an inter-house school competition for the modern age - an esports part of the new esports qualification that he and another six other pupils are studying at Selkirk, one of the schools piloting the new qualification. The new National Progression Award (NPA) in Esports, which is equivalent to a Higher, provides an overview of the esports serves as a starting point for learners who want a career in the sector, which is growing rapidly with increasing employment opportunities."When I first looked at the course I thought this was a strange course I wouldn't expect to be at school," says Alfie. "But then I thought wow, this is really exciting."Alfie stresses that the course is not about playing video games. It's about studying the background and history of esports - and exploring the opportunities of the multi-billion pound global didn't realise that people could make a career out of gaming."We've learnt about the history and the business side and how people make money," he said."I didn't know this but you can be a nutritionist for players."The tournament is the culmination of months of work and planning for the team."It's been quite hard because you had to find out where can you plug stuff in, how do you actually get enough people that actually want to play and getting all the equipment sorted," Alfie said. Paul Graham, an education development officer who specialises in digital education with Scottish Borders Council, said there was some initial scepticism when the esports qualification was introduced."I think when you talk about esports and you start to highlight the strengths and the skills that young people can get from it, the passion they have for it and the learning that can come alongside, it starts to really make sense when they start to hear what those opportunities are," he added that actually playing games is a very small aspect of the course which he says teaches the pupils a range of valuable and transferable skills in a bid to inspire the next generation of creators, developers, and entrepreneurs."We want every school to have an opportunity to run these esport courses and gain qualifications up to the equivalent of Higher but then also allow young people to take that to college courses and then on to university level if they want to," he said. The tournament is a big success as far as the teachers are concerned, the deafening noise from the cheering crowds of pupils a sure sign they enjoyed in the classroom the esports class is already breaking down what worked and what didn't, how they could have made more from it and how to tackle the technology problems they de-brief is being led by the course teacher Richard Willan, the principal teacher of creativity and enterprise at Selkirk High. He is impressed with the effort the pupils have put into the tournament, not least the amount of their own time out of school."This course actually puts all the learning into context." he said."They develop research skills, teamwork skills, interpersonal skills, but also the literacy and numeracy skills that they have had to use, all based within the esports context, on something they are actually engaged and interested in. "The learning has just happened. From a teaching perspective it's amazing because esports is just the vehicle for them learning and its a nice neat package and that means that we are not having to push the learning, they are driving it themselves."After the tournament Alfie says he has loved being part of the organising team, he's already full of ideas for another one. He has clearly taken a lot from the event and says he is definitely interested in a career in esports or the gaming when I asked him would he rather a career in esports or a rugby cap for Scotland, there's not a moment's hesitation: "Play for Scotland".
