Anthropic's Claude Fable 5 is going to be a big hit with the web's vibe coders.

Cities are dynamic, not static grids, and urbanization is a "spiky," cyclical, and asynchronous process.

Five peer-reviewed papers update the design and model its expected output.

“Ghost-Sender" is the result of a widespread misconfiguration, according to researchers, and evidence indicates it's being actively abused in the wild.

Netflix's response: "Absurd."

By 2050, the US median home-sale price could be $1 million, according to a forecast from NAR's economist. d3sign/Getty Images Data from NAR shows the median sale price of an existing home was $429,300 in June. The median single-family home sale price could hit $1 million by 2050, according to an economist. The climb to $1 million will be slow, so buyers should expect modest price gains for now, they said. I know it, and you know it — buying a home has never felt this unaffordable. For years, prospective buyers have faced a punishing mix of too few homes for sale, high mortgage rates, and persistently high home prices. Those pressures have pushed homeownership out of reach for many Americans. Unfortunately, affordability could get even worse in the years to come. At an economic forum during the National Association of Real Estate Editors conference in June, National Association of Realtors chief economist Lawrence Yun told a room full of journalists that he expects the median single-family home sale price to reach $1 million by 2050. For comparison, the median sale price of an existing home was $429,300 in June, according to NAR data. Yun told Business Insider that while his 2050 forecast is speculative, it is based on steady annual home-price growth of 3% to 4%. "It's worth looking back: for example, back in 1990, the median national home price was barely above $100,000," Yun said. "At that time, people would not have envisioned that today's home prices could be $400,000." He pointed to San Francisco's housing market as another example. "In San Francisco, a million-dollar home is very common today, but back in 1990, the median home price in San Francisco was about $250,000, and I'm sure no one was thinking about a million," Yun added. Yun's forecast is sobering, especially as some parts of the real estate market appear to be normalizing. Home-price growth has slowed in formerly hot markets like Austin and San Diego, while new home construction in smaller markets like Kaufman, Texas, and Harnett, North Carolina, has helped boost sales. Yun said those cities are outliers in a national real estate market still defined by climbing prices and soft demand. "Places like Austin, or any local market, will deviate from national trends for some time. But Austin's price decline is pretty much due to the fact that it had a spectacular run-up," he said. "One wonders, after a little pause in Austin, whether it will begin to show that superstar status again." Yun said it will take years of steady gains for the US median single-family home sale price to reach $1 million. For now, he expects prices to rise only moderately. "In the short term, I think home prices will see a very minimal gain this year — maybe 1%, 2%, or 3% nationwide," he said. "That would be below consumer price inflation and below income growth, since incomes are rising around 3% to 4%. I think home-price growth coming in slightly below income growth is a healthy thing." Read the original article on Business Insider

A huge change is coming to the world’s booming artificial intelligence (AI) sector. Starting with Elon Musk’s SpaceX, with OpenAI and Anthropic preparing to follow, all three private companies are set to sell shares of their stock to the general public for the first time. These are what’s known as initial public offerings (IPOs). SpaceX – the first of them to launch this Friday, June 12 – expects to raise $US75 billion from selling just 4% of the company’s shares. Musk is already the world’s richest man, worth around US$800 billion. He owns around 42% of SpaceX now, plus options to buy more shares at a fraction of the US$135 a share price ordinary investors are being asked to pay. Given his existing wealth, after this Friday’s listing Musk looks likely to become the world’s first trillionaire. Together, these three companies are valued at almost $US4 trillion and are expected to raise a record-breaking $US200 billion, despite well-founded concerns that big AI stocks are now hugely overvalued. While most of the news coverage has focused on the money involved, there’s actually another side to these sales that could be a big deal in the longer run. At a time when everyone from the Pope to people from all walks of life worldwide are concerned about AI’s growing role in our lives, these stock exchange listings have the potential to finally bring some extra transparency to the inner workings of the AI giants. Why SpaceX, OpenAI and Anthropic matter to you Once these companies list, hundreds of millions of investors around the world will be exposed to these companies. That could be directly, if you buy these stocks, or else through index funds, which hold shares on behalf of investors – including big retirement and superannuation funds. Even for those who don’t consider themselves investors, these three share offerings could easily affect your savings too. Here’s what we know about these IPOs so far. SpaceX’s most recent June 3 filing amendment with the United States Securities and Exchange Commission added a notable new line. It said SpaceX “may issue a significant amount of equity in connection with future transactions”. US business outlet Fortune is reading this as a signal for a possible future Tesla merger, bringing another of Musk’s companies into the fold. That could be the biggest merger in history. Read more: Switzerland’s entire GDP: visualising Elon Musk’s record-breaking pay deal Investors can sue over failures to disclose Once publicly listed, the AI labs of SpaceX – xAI – as well as Anthropic and Open AI would be subject to public market scrutiny for the first time. This would push these companies to disclose more AI risks than they have had to as private companies – or risk being sued for misleading investors. US securities laws are among the most enforceable in the world. Under US law, investors can sue a company for securities fraud if it fails to disclose a risk that later materialises. One regulation commonly used in securities fraud lawsuits is Rule 10b-5 under the Securities Exchange Act of 1934. This has been successfully used in the past many times. For example, Bank of America paid US$2.43 billion settle a lawsuit related to its purchase of investment bank Merrill Lynch in 2008. Countrywide Financial paid US$600 million for failing to disclose the mounting risks of its subprime mortgage business. Only last month, the International Monetary Fund warned “financial stability risks mount as artificial intelligence fuels cyberattacks”, pointing out: Anthropic’s recent controlled release of its Claude Mythos Preview, an advanced AI model with exceptional cyber capabilities, underscored how quickly risks are increasing […] This foreshadows how fast‑moving, AI‑driven cyber risks could destabilize the financial system if not managed carefully. There are good reasons to be concerned about the increasing dominance of tech companies and what happens to economies around the world if the AI share bubble bursts. Having more of the biggest AI companies forced into greater disclosure would offer one silver lining amid those AI fears. AI and chip stocks have been surging in 2026. What happens if the AI bubble bursts? What difference could public disclosure make? Just as an example, let’s suppose Anthropic accidentally leaked its Claude Mythos source code (like a leak that actually happened earlier this year). Then let’s say North Korean hackers used that code to hack into US government systems. If that happened when Anthropic was a public company, its share price would very likely fall in response. Investors could then sue Anthropic for failing to disclose the risk of code leak, which later caused the share price to fall. This mechanism has its limitations: it only works if AI harms are eventually reflected in stock prices of Anthropic. In other words, the mechanism only protects the general public from AI risks indirectly – though protecting Anthropic’s investors first. Musk’s SpaceX is shaping up as the biggest IPO on record. It’s also bending the rules to do so How much more accountability should we expect? The market is meant to incorporate all public information to arrive at the fair price of a public company. In doing so, market listings should make it easier for investors to police AI safety. After all, it’s in investors’ interests to not drive humanity to the verge of collapse. But is the market delivering on this function so far with AI? So far, you’d have to say it’s not. For instance, the world’s second-largest stock exchange, the New York-based Nasdaq, controversially changed its own rules for SpaceX to join its Nasdaq 100 index after just 15 trading days, not the usual three months. But perhaps there is still hope that investors’ own desire to survive AI will make them push companies to manage AI’s risks more responsibly. Will it be enough? Probably not on its own. The risks most people worry about with AI – diffuse, slow-moving, hard to pin to a single quarter – may never register clearly in an earnings report. But more disclosure is better than less. And more disclosure is exactly what these listings will finally force. Marta Khomyn does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

Discussion | Link

Joe Rossignol / MacRumors: Apple says 79% of all iPhones and 86% of iPhones released in the last four years were running iOS 26 as of June 7, vs. 82% and 88% for iOS 18 before WWDC 2025  —  Apple has shared updated iOS 26 and iPadOS 26 adoption figures, revealing how many iPhones and iPads were running those software versions …

The attacks stemmed from a GitHub account that was also compromised in a previous Miasmi attack on Microsoft last month.

Taniquewa Brewster, 41, is a former basic income participant in Austin, Texas. Courtesy of Taniquewa Brewster Taniquewa Brewster received $1,000 a month for one year from an Austin basic income program. She landed job certifications and stable housing, but is worried about emergency expenses. US cities have run hundreds of pilots offering no-strings-attached cash to low-income families. Taniquewa Brewster is an assistant property manager, doula, student, advocate, and proud mom of six. The 41-year-old is also a former guaranteed basic income participant in her hometown of Austin, Texas. She was one of 135 low-income households that received $1,000 a month for a year, from September 2022 to August 2023. The money helped her cover bills, build savings, and land a more stable career. Business Insider first spoke with Brewster in the fall of 2024, about a year after her cash payments ended. She's now feeling more confident in her family's future — even if unexpected bills remain a stressor. "I've been able to keep my children from having to work to contribute to the household," she said in June. "They can do whatever they want to do with their money, and that's a big difference for me — breaking that generational curse of poverty." She said that fear is no longer her dominant emotion. Basic income gained momentum in the early 2020s as a solution for poverty and an increasingly AI-dominated job market. The no-strings-attached cash model has been tried hundreds of times over the past decade, championed by lawmakers, local advocates, and some big-name tech leaders. Business Insider has heard from single parents who used the money for groceries and families who secured apartments after years of instability. Much of the reporting and data about basic income follows participants during and immediately after their programs. What's less documented is how families fare in the years after their cash payments end. Brewer said she's still seeing a benefit. Do you have a story to share? Reach out to this reporter at allisonkelly@buinessinsider.com. Brewster feels more financially confident, but worries about emergency expenses Austin's guaranteed basic income program was a collaboration between the city and the nonprofit UpTogether. Brewster qualified because she was experiencing housing insecurity and earned less than 60% of the area median family income, which was slightly over $64,000 annually for her household size. Brewster is a single mother who supports her children — now between infancy and college age — alone. Prior to GBI, she was patching together low-paying part-time jobs and struggling to pay bills. She said the financial strain prevented her from getting the training she needed for a higher-paying real estate job with predictable hours. "I was working as hard as I could, working overtime," Brewster said in 2024. "That's keeping me away from my family. That's keeping me away from things that I can be doing — even going back to school — because I have to work so hard to just cover rent." With the extra $1,000 a month, Brewster said she was able to get the certifications she needed to be promoted as a leasing agent and now assistant property manager. She also completed her doula training and has since gone back to school to study nonprofit management. She remains a vocal advocate for basic income in Texas. Building a career that both supports her family and allows her to support others is a dream, she said. She's currently adopting her youngest daughter, whom she's been fostering, and her family is preparing to move to a bigger apartment. Data from the Austin pilot shows that the majority of participants were better able to afford basic necessities and pay off debt during and directly after the pilot. "It didn't just help me financially, it gave me confidence in myself," she said. "It was a hand up and not a handout." She said the payments didn't fix all of her affordability concerns. She had an unexpected illness last year that required expensive medications and hospital care. She has some savings, but worries about what will happen if her family runs into more emergency expenses. Like other GBI pilots, Brewster's program was designed to provide short-term financial support. Austin's program stopped recently due to a funding lapse and limited political backing. The US has yet to try basic income for more than a few years at a time, largely because governments and philanthropists say they can't fund it long-term. Some tech leaders who previously championed basic income for displaced workers now say the impact doesn't justify the investment. Critics of basic income also say cash payments discourage participants from working to earn a paycheck, and that the programs require too much capital to operate alongside other safety net programs, like SNAP and Medicaid. But, for Brewster, even a year of no-strings-attached cash was life-changing. "It felt like people were actually asking me what my needs were, and how they could help me meet those needs," she said. "There's something comforting about someone actually caring enough to ask." Read the original article on Business Insider

New frontier model refuses cybersecurity, biology, and chemistry queries.

<blockquote cite="https://twitter.com/karpathy/status/2064409694761054332"><p>I feel a lot of things changing as working software increasingly comes out on a tap. The Jevon's paradox kicks in and I feel my own demand for software growing substantially. You can ask for anything - explainers, visualizers, dashboards, bespoke single-use apps (e.g. a full wandb that is hyper-specific just for your project), you can 10X your test suite, auto-optimize code, run giant research projects with custom HTML for the results, anything! "Free your mind" (Matrix ref).</p></blockquote> <p class="cite">&mdash; <a href="https://twitter.com/karpathy/status/2064409694761054332">Andrej Karpathy</a>, on Claude Fable 5</p> <p>Tags: <a href="https://simonwillison.net/tags/andrej-karpathy">andrej-karpathy</a>, <a href="https://simonwillison.net/tags/jevons-paradox">jevons-paradox</a>, <a href="https://simonwillison.net/tags/anthropic">anthropic</a>, <a href="https://simonwillison.net/tags/generative-ai">generative-ai</a>, <a href="https://simonwillison.net/tags/ai">ai</a>, <a href="https://simonwillison.net/tags/llms">llms</a>, <a href="https://simonwillison.net/tags/claude-mythos">claude-mythos</a></p>

As AI infrastructure scales, enterprise expectations for operational maturity are increasing. Organizations expect these systems to be provisionable,... As AI infrastructure scales, enterprise expectations for operational maturity are increasing. Organizations expect these systems to be provisionable, observable, secure, and manageable at scale—the same standard applied to all critical infrastructure. The moment an AI system moves from development into enterprise deployment, that operational foundation is essential. NVIDIA DGX Spark and… Source

Discussion | Link

Voice translations preserve speaker's tone, pacing, pitch—with SynthID watermarks for security.

Milana Vinn / Reuters: Sources: Canadian payments company Nuvei is in advanced talks to acquire cross-border payments company Payoneer Global for about $2.7B  —  Canadian payments firm Nuvei is in advanced talks to acquire cross-border payments company Payoneer Global (PAYO.O) for about $2.7 billion, according to two people familiar with the matter.

"Artemis III will be an extraordinary demonstration of what is possible."

Converting a quantized checkpoint into an NVIDIA TensorRT engine bridges the gap between model optimization and production deployment, enabling faster... Converting a quantized checkpoint into an NVIDIA TensorRT engine bridges the gap between model optimization and production deployment, enabling faster inference, higher throughput, and more efficient GPU utilization at scale. In a previous post, we produced a high-quality FP8-quantized Contrastive Language-Image Pretraining (CLIP) checkpoint with NVIDIA TensorRT Model Optimizer. Source