A (0:01)

This episode is brought to you by State Farm. You know those friends who support your preference for podcasts over music on road trips? That's the energy State Farm brings to insurance. With over 19,000 local agents, they help you find the coverage that fits your needs so you can spend less time worrying about insurance and more time enjoying the ride. Download the State Farm app or go online@statefarm.com like a good neighbor, State Farm is there.

B (0:31)

Welcome to the Techbrew Ride Home for Thursday, April 9th, 2026. I'm Brian McCullough. Today Meta releases its first model in the Great AI Horse Race Catch Up Race. OpenAI is expecting to make as much money from ads as the biggest social media companies. Meta doesn't want you to see ads encouraging you to join lawsuits against them, and Amazon's big starlink competitor is coming as soon as this summer. Here's what you missed today in the world of tech. Well, they're trying to stay in the race here. Meta has released Muse Spark, the first model from Meta superintelligence Labs under the direction of Alexander Wang to quote power a smarter and faster Meta AI across all of Meta's products. Quoting the ft, Mark Zuckerberg signaled earlier this year that the release of these initial AI models would not yet match the most advanced systems from rivals. But the company said Muse Spark would power a smarter and faster version of Meta AI, its virtual assistant. It added that the model would enable more personalized and visual responses, drawing on content shared across Instagram, Facebook and threads. Expect richer, more visual results with reels, photos and posts woven directly into your answers, with credit back to the content creators, Meta said. Meta's shares were up 8% on Wednesday following news of the model amid a broader market rally. It also said its new model had strong applications in healthcare. Meta said it had worked with more than 1,000 doctors to train the model to generate more detailed responses on top, such as nutrition and exercise. Using the model, Meta AI will also offer a shopping mode to help users compare prices. Unlike Meta's previous Llama models, which have been open or freely available for developers to tweak, musespark is a smaller, closed model. The company intends to offer a private preview to select partners. Meta said it hoped to open source future versions. The release offers an early indication of how Mark Zuckerberg plans to incorporate Meta's social media content into its chatbot. As part of his promise to develop what he calls personal superintelligence, Musespark beat leading models from Google, OpenAI and Anthropic and Select Benchmarks on reasoning and multimodal capabilities, according to evaluations shared by Meta. But it acknowledged that the model was not yet cutting edge in certain areas, adding, we continue to invest in areas with current performance gaps, specifically long horizon agentic systems and coding workflows. Zuckerberg said on Wednesday that Meta planned to release increasingly advanced models that push the frontier of intelligence and capabilities, including new open source models. We are building products that don't just answer your questions, but act as agents that do things for you, he added in a post on Threads With Muse Spark, Meta appears to lean into niche capabilities such as health, arguing that it is one of the top reasons people turn to AI. Rival OpenAI in January launched ChatGPT Health, a specialized feature that allows users to connect their medical records and fitness apps to its AI to get personalized health advice. Meta has also been working to improve the model's multimodal perception, such as recognizing objects and photos. Zuckerberg last year said Meta was doubling down on its investment in AI glasses, arguing wearable devices are vital to his bet on superintelligence. When Meta AI powered by Muse Spark comes to our AI glasses, the assistant will be able to better see and understand the world around you, the company said. End quote. Meta also said they are opening a private API preview for Muse Spark to select partners. I think I mentioned that and plan to offer paid API access to a wider audience later. Meanwhile, I found this from Ethan Mollick on X interesting in terms of where we are right now in the broader AI horse race. Quoting a Twitter thread from Malik so now we have a pretty good picture of the state of the Frontier AI model makers. US closed source models continue to lead Google, OpenAI and Anthropic stand well ahead of the pack and may have signs of recursive self improvement. AI has fallen from frontier status for now, though promises to return shortly. Meta Re entered the space today with a not quite frontier closed source model, but an approach that suggests that they might be back in the race. All the other US players seem far behind on the Chinese model front. Alibaba with Quen, Moonshot with Kimmy, Minimax, Xiaomi with Mimo, Deepseek and Z all still appear to be very much in the race, though the best Chinese models are still seven to nine months behind released US closed source models. For some of these players, especially Xiaomi, and Alibaba, their commitment to open weights appear to be slipping. Lots of questions still remain. Is recursive self improvement indeed happening at the Big three labs? How long can the exponential of LLM ability gain? Last, how much do Chinese models rely on distillation and can they keep pace given chip constraints? Will there continue to be Frontier openweight models? The US Frontier Labs have all walked away from openweights. They continue to occasionally release excellent open models like Gemma 4, etc. But they are smaller models that are not competitive with their closed weight model kin. So all eyes are on Chinese AI labs for open models. More signs of where investors think Anthropic stands in the AI horse race Quoting Bloomberg Anthropic employees have sold some equity to investors, wrapping up a secondary share sale that started earlier this year, according to people familiar with the matter. But some investors weren't able to pick up as many shares as they planned because of the limited number that employees were willing to sell. The tender offer took place at the same value as the company's most recent fundraising in February, said the people, who asked not to be identified discussing private information. The company was valued at $350 billion in its latest deal, not including the $30 billion it raised. The total value of the share sale, which closed last week, could not be learned, but it fell short of the amount that investors had lined up, which was as much as $6 billion. Some of the people said current and former employees wanted to hold more of their shares ahead of Anthropic's upcoming initial public offering, expected as soon as this year. Some investors were able to get their full allocation in the deal, while others were only able to deploy some of the capital that they had set aside for the tender offer. The smaller than expected transaction suggests that employees are optimistic about the company's prospects as its annualized revenue climbs, one of the people said. Last month, the company surpassed $19 billion in annualized run rate revenue. By April, Anthropic announced it had surpassed $30 billion in run rate revenue. This is not exactly the same thing as that, but OpenAI does seem to be experiencing explosive growth in one aspect of their business. Quoting the information OpenAI's advertising effort is so new that the company is still working out the kinks in it. That hasn't stopped it from setting ambitious aims for the business, which it expects to become the largest driver of its revenue by the end of the decade. OpenAI expects advertising to generate about $2.4 billion in revenue this year and to quadruple next year to nearly $11 billion, according to financial forecasts from the first quarter. A year ago, the company forecast it would generate $1.6 billion in revenue this year and $5.9 billion next year from users who didn't pay for subscriptions, an amount the company likely expected to derive from advertising and e Commerce. In 2030, OpenAI expects ADS to generate about $102 billion, or 36% of its total revenue for that year. A year ago, it forecast non paying users would only generate $26.5 billion in sales that year. If it can succeed in hitting ad sales of $102 billion, that figure would be about half of what Meta platforms generated from advertising last year. Advertising sales are likely to increase in part through the growth of OpenAI's audience, but forecasts for higher revenue per user suggest it also expects to charge more for ads, show ads to its users more frequently, or both. The ChatGPT maker projects the average revenue per user for its ads business to surge to about $12 next year from nearly $3.5 this year, according to the recent forecasts. By 2030, OpenAI expects ARPU to rise to nearly $60. That's annual revenue per user. That's four times higher than the company's early 2025 forec for its 2030 average revenue per non paying user. Meta's average revenue per user was about $57 last year. The forecasts also highlight how dependent OpenAI is on keeping users enthralled with its AI. Weekly active users have topped 920 million, short of its projection of 1 billion by the end of last year. Most people don't pay for ChatGPT subscriptions, but the personalized conversations they have with the chatbot represent an intriguing new opportunity for advertisers. It's not clear how many weekly users anthropic has for OpenAI, figuring out the ads model will be essential in order to commercialize that ad's business, said Tomasz Tunguz, founder of software investor Diri Ventures, on TITV Wednesday. It's about a $500 billion ad market today. What fraction of that will go to AI? It'll be pretty significant.