Tech Brew Ride Home - GTA Now

Episode Date: June 24, 2026

Rockstar finally priced GTA VI at $79.99 and set a November 19 release, with preorders tonight. OpenAI and Broadcom unveiled their Jalapeño inference chip. Meta got caught building a prediction-markets app called Arena, and Superhuman snapped up AI-detector GPTZero. Rockstar sets the release date for GTA VI for November 19 and says it will cost $79.99, or $99.99 for the Ultimate Edition; preorders start at midnight tonight (The Verge) Grand Theft Auto 6 Physical Copies Won't Include a Disc, Will Just Be a Code in a Box (IGN) OpenAI and Broadcom unveil Jalapeño, an LLM-optimized inference chip developed from design to manufacturing tape-out in nine months, aided by OpenAI's models (OpenAI) OpenAI and Broadcom Unveil AI Chip to Run Models Faster, Cheaper (Bloomberg) Sources: Meta is building a standalone prediction markets app internally called Arena, which would probably use video game-like points instead of money wagers (The New York Times) Superhuman acquires AI detection startup GPTZero, which has 19M+ registered users and $30M in annual recurring revenue; PitchBook: GPTZero is valued at $88M+ (Business Insider) How AI Customers Are Lowering Their Anthropic and OpenAI Bills (The Information) Subscribe to the ad-free feed. Learn more about your ad choices. Visit megaphone.fm/adchoices

Welcome to the Tech Brew Ride Home for Wednesday, June 24th, 2026. I'm Brian McCullough today. Rockstar finally priced GTA6 and named the release date with pre-orders starting tonight. Open AI at Broadcom unveiled their jalapeno inference chip. Mata got caught building a prediction markets app called Arena and Superhuman snapped-up AI detector GPT-Zero. Here's what you missed today in the world of tech. I'm getting to the age where I might pull a muscle just by thinking about stretching. That's why Blueprints' longevity mix has been such a a game changer. Blueprint was started by Brian Johnson, more popularly known as the Don't Die guy. His latest endeavor in Living Longer and Aging Better led him to create Blueprint. With just one scoop,

you'll get support for energy, cognitive performance, mood, focus, sleep, cellular resilience, and healthy aging. Science-backed, precision dose, no BS. For a limited time only, our listeners get 20% off, plus free shipping at Blueprint.briamjohnjohn.com by using code TBRH at checkout. That's code TBRH at Blueprint.Briennyson.com for 20% off. And after you purchase, they will ask you where you heard about them. Please support our show and tell them our show sent you. At long last, GTA6 Order Day is here. Quoting the Verge, we finally have a price for Grand Theft Auto 6, $79.99. That gets you the standard edition of the game, while the ultimate edition will set you back $99.99. cents. Pre-orders start at midnight tonight, local time, when you'll be able to reserve a copy on

PS5 or Xbox Series X-slash-S. You'll be able to order either digital or physical editions of the game, though Rockstar has confirmed that even the physical editions contain a download code inside the box. More on that in a second. The upside is that they'll ship from November 12, allowing buyers of either addition to preload the game ahead of its full release on November 19th. Anyone who pre-orders or purchases the game before November 20th will receive the vintage vice City Pack, a throwback to 2002's GTA Vice City that includes a 55 vapid stanier sedan and garage, outfits and hairstyles, and a weapon skin. Digital pre-orders also net a free month of Rockstar's GTA Plus subscription. As for the Ultimate Edition, that includes digital extras unlocked throughout

the game's story, including vehicles, weapons, outfits, and tattoos, and exclusive in-game stores. There's been a lot of speculation about GTA-6's price leading up to today's announcement, gaming in general has steadily gotten pricier over the past year with both console and game prices going up like Nintendo's move to nudge some Switch 2 game prices up to $69 or even $79. It's been a long wait for GTA 6. It will launch over a decade after GTA 5 came out and its release date has been bumped back a couple of times. It's expected to be the biggest game of the year, so much so that other games are planning

their own launches to avoid GTA 6 entirely, end quote. back to that story about this being another nail in the coffin for physical media. Quoting IGN. Rockstar will sell physical copies of Grand Theft Auto 6, though you'll only get a box containing a digital download code. This morning, Rockstar announced that it will finally begin GTA6 pre-orders at midnight tonight, with various options available. Alongside, a digital pre-order fans will also be able to buy GTA6 in a box, though no disc will be included inside. On the upside, physical copies will be available from November 12th in advance of the game's actual launch on November 19th. This earlier date is also when GTA6 preloading will become available allowing anyone with a physical copy or digital pre-order to get the game downloading.

While discless physical copies have become more common in recent years, it is notable that such a big mass market game, likely this console generation's biggest most market game, is ditching physical media altogether. but in a world where even Nintendo is ditching physical media for many of its own exclusives such as Pokemon Pocopia, it is perhaps not too surprising. Notably, the GTA6 Ultimate Edition will include exclusive vehicles, weapons, and outfits as well as additional locations and activities that it appears will be absent from the base version of the game, a cunning move by Rockstar to offer two entry price points, but Nudge committed fans to spend $100. And that's all before any new version of GTA Online arrives too. According to Rockstar, GTA6 features a single-plice player experience launching on November 19th to PlayStation 5 and Xbox Series X-slash-S, copies of the game

purchased before November 20th, will at least include an additional vintage Vice City Pack with more cosmetics and cars, though physical addition customers should note that this is only guaranteed while supplies last. Digital pre-orders also get you a free month of GTA Plus, end quote. OpenAI unveiled its first custom artificial intelligence chip developed in partnership with Broadcom, part of a bid by the chat GPT maker. to gain an edge by tailoring hardware to better run its AI products. Quoting Bloomberg. The company said Wednesday that OpenAI has received the first samples of the chip called Halapeno

and is testing how the silicon handles running AI workloads. So far, the accelerator is showing cost savings of roughly 50% compared with typical AI graphics processing units. Broadcom chief executive officer Hock Tan said in an interview, the finalized chips will be integrated into large data centers from Open AI backer, Microsoft, and other partners starting later this year. Tan expects OpenAI and Broadcom should be able to exceed his prior forecast for deploying 1.3 gigawatts worth of chips next year. We like to think we can do better because there is a lot of demand, he said.

Though Open AI continues to rely heavily on chips from market leader, Nvidia, the startup has worked to broaden its mix of suppliers to meet surging demand for its AI services. Open AI has struck multi-billion-dollar deals with chipmakers such as advanced microdevices and cerebrus. In October, OpenAI announced that it would part-eastern. with Broadcom to design accelerators tailored to work best with its artificial intelligence models. Open AI plans to spend tens of billions of dollars on Broadcom chips, Bloomberg News has reported, adding to the unprofitable startup's massive outlays on physical infrastructure to support AI.

Open AI's head of hardware, Richard Ho, declined to say how the company will pay for the chips, noting that financing arrangements would be finalized when the full order of processors is complete. Tan declined to comment, but reiterated that Broadcom has set up a chip financing vehicle with Apollo Global Management, and Blackstone that will also aid OpenAI. The new chips were developed from scratch in record time, according to OpenAI and Broadcom. OpenAI said it is still assessing the test versions, but early data shows Halapeno is capable of delivering performance per watt of energy that's substantially better than current state of the art. The chip was designed for large language models or LLMs and guided by the artificial intelligence startup's approach for inference,

a stage that involves running AI tasks, host said. Still, he said, it is a very general purpose device that will address future LLM innovations. Open AI wants to keep its options open as to whether it would let other model makers use its chips, Ho said. Within the tech industry, some AI chips are more powerful for doing many tasks at once, while others are focused on providing a single answer to a user as fast as possible. Open AI said its goal is to combine both the power of leading AI chips with speed that's closer to specialized hardware for running models faster. The chips are meant to perform better by reducing the amount of data that must move around. The hardware is also designed around the usage patterns of computing, memory, and networking

gear that are most important for top-of-the-line AI models. The two companies have a roadmap for future chips with the next version plan for 2028, and then annually after that, Tan said, while Halapino is focused on inference, OpenAI may consider other workloads with future chip generations, end quote. Meta is apparently building a standalone prediction markets app internally called Arena, which would probably use video game-like points instead of actual money wagers. Quoting the times, Polymarket and Kalshi, prediction markets where users can bet on outcomes as varied as the Super Bowl and the length of the state of the union address, have been some of

the fastest growing destinations on the internet. Mark Zuckerberg has noticed, and he wants in on the action. Mr. Zuckerberg, the chief executive of META, recently dispatched a small team at his company to create a smartphone app similar to Polymarket and Kalshi. Two employees with knowledge of the matter said users would not wager money, and the app would probably rely on a video game like point system. Instead, one person said, though the company had not ruled out the eventual use of real money betting. The app is internally referred to as arena and would function independently from meta's social networking apps, which include Facebook, Instagram, WhatsApp, and Messenger, said the employees who spoke on the condition of anonymity. to discuss confidential plans. Meta aims to grow the app by leveraging its large social networking audiences and directing them toward using it, they said. The effort which Insiders characterized as experimental,

but a top priority is part of a broader push by Mr. Zuckerberg to create new types of apps based on emerging social behavior online. More than 3.56 billion people visit one or more of meta's apps every day, an amount that has raised questions about whether those platforms have reached a saturation point. Arena is one of a handful of apps that Meta is trying out. Others include one called Meta Photos, another standalone app, which would create new types of media using artificial intelligence, the employees said. This is not the first time meta has experimented with prediction markets. In 2020, it released Forecast a crowdsource prediction market app that prompted people to make guesses about the world in the early days of the COVID-19

pandemic. The app was positioned as a way to share crowdsource knowledge. It used a point system to make predictions about the future, and Meta shuttered the app in 2022. Since the app, then, prediction markets have exploded into a cultural phenomenon featured during major sports events and in the Golden Globes Telecast. In 2025, Kalshian Polymarket drew a combined $50 billion in online trades. This year, the total has already surpassed $130 billion, end quote. As a podcast called The Ride Home, we are very much anti-crash. But keeping your energy up without coffee or nicotine, that's no small feat. Thankfully, ultra-pouches, plant-powered, pouches are designed to sharpen your mind and lift your energy all without caffeine, nicotine,

or the crash. Infinity PX, L-theonine, Alpha GPC, and vitamin B6 and B-12 help you achieve a calm, steady flow-state focus. Plus, they come in all sorts of flavors like cool mint, wintergreen, tropical, watermelon. New customers get 15% off with code TBRH at take ultra.com. That's take ultra.com. TBRH. When inspiration strikes, you're usually not sitting at your desk ready to take notes. And honestly, some ideas are easier to say out loud than to type into your phone.

That's where Plod Notepin S comes in. It's a small, wearable AI-powered device about the size of a AA battery that lets you capture ideas and conversations completely hands-free. So in those in-between moments, when pulling out your phone just feels like too much, you can just tap once and keep going. Plod automatically turns what you record into searchable notes, summaries, and mind maps you can actually use later. Plaud's also designed with privacy and mind with enterprise-grade standards including SOC2, HIPAA, and GDPR-CMPLINC-T-R-H-T-R-H for 10% off. Gee, we were just talking about this yesterday, quoting Business Insider, as A-U's code.

AI-generated content storms the internet. Productivity company Superhuman is acquiring GPT-Zero, an AI-detection startup co-founded by Edward Tian and Ux Sioux in 2023. GPT-Zero, which Tian, now 26, built as a senior at Princeton, went viral after its launch and has since evolved to include additional tools to detect hallucinations and AI-generated content in social media feeds. The company's declined to share the financial terms of the deal. GPT-Zero is valued at over $88 million, according to Pitchbook,

and it's backed by investors like Uncourt Capital, Neo Footwork, and Jack Altman. GPTZero has grown rapidly since its launch three years ago, Tien said. He added that it surpassed 19 million registered users and $30 million in annual recurring revenue. As part of the acquisition, Tien and two will join Superhuman to lead a team focused on authenticity. GPT Zero's 30 employees will also join Superhuman. Superhuman CEO, Chishir Marota said, When you're buying a business like this, the people come first, Marota said of GPT Zero's co-founders. Superhuman was formerly known as Grammarly and changed its name after acquiring

Superhuman, an email app popular across Silicon Valley and among venture capitalists. It now offers a variety of productivity tools, and its acquisition of GPTZero aims to strengthen its existing authenticity tools, including its Grammally AI detector and an authorship product that will help writers verify their work. GPTZero will soon be accessible within Superhuman Go, an AI assistant that operates across websites and apps. GPT-Zero will also continue to operate as a standalone product. Deciphering which content is and isn't AI-generated is especially critical in education, Marota said.

And demand is also coming from professional fields such as consulting, recruiting, and journalism. Education accounts for roughly a third of the more than $700 million in annual revenue of Superhumans flagship writing assistant grammarly, he said, while professional users generate the remainder. GPTZero started with the mission of preserving what's human, Tian said, now we need to preserve critical thinking. GPT0 marks Superhumans' fourth major acquisition, Marota said. Previously, the company acquired the Productivity Assistant Coda,

the email app, Superhuman, for which it is now named, and the AI spreadsheet tool rose, end quote. Finally today, what if you could get your AI 28 times cheaper? We've been talking about this possibly happening, and the information says it is actively happening. As prices for Anthropics and Open AIs flagship products soar, some large customers are using cheaper AI models from those companies' menus as well as from other providers.

Ensemble Health Partners, a provider of software for hospitals that plans to spend up to $100 million on AI this year, says it's had success switching to an Open AI model that's $1.20th as expensive as the company's more advanced models. The AI powers a tool for writing appeal letters to insurance firms that won't reimburse Ensemble's hospital clients for care they provide. Ensemble sends about 15,000 such letters every month, according to Grant Viesi, its chief technology officer, so switching to the lower-cost model will result in savings of nearly $700,000 per year. Many Anthropic and OpenEye customers are willing to stomach the rising costs of the two companies' products, but some are

hitting the brakes after blowing through their budgets. Some are developing or buying software known as routers that can automatically pick lower cost models depending on the complexity of the task. Others such as meta platforms and Uber are limiting how much AI employees can use in a given week or month after previously allowing nearly unlimited use. And some companies are also controlling which types of employees get access to the most advanced and expensive models and which ones don't. Sometimes reducing AI costs can be as simple as telling an AI model to do less thinking. Enterprise Software firm UiPath says it's been using prompt engineering to minimize how much models spend warming up or thinking before doing a task, especially tasks they've done before,

which led to more than 90% cost savings for some tasks. But nothing saves money quite like finding that an open source or cheap model can handle a company's immediate AI needs. Finn, which sells AI-powered customer support agents, began switching from Anthropics' lower-cost sonnet models in favor of developing its own custom AI using open-weight models more than a year ago and has accelerated that transition in recent months, said Faragal-Reed. chief AI officer. He said the change has saved the startup millions of dollars annually and has helped it improve its profit margins significantly, though he declined to disclose what open source models Finn uses. Salesforce last week agreed to buy Finn for $3.6 billion and also could be

interested in reducing its hefty, Anthropic, and Open AI bills. It's hugely significant to the point that anyone running an application company should be trying to do something like this, Reid said. Since the U.S. government earlier this month banned foreign access to Anthropics' latest AI model, foreign enterprises and government agencies have been speeding up plans to use models from non-U.S. sources, according to Cohere, a Canadian AI firm that provides access to small and open-source models. Since the U.S. restrictions went in place, Cohear tripled its internal projection for annualized revenue in 2027, said spokesperson Kyle Lastovica, without disclosing the figure. Coher's annualized revenue, typically a measure based on the prior month's revenue multiplied by

12 was $240 million at the end of 2025, according to a person with knowledge of its financials. The surge in serious inbound interest from large buyers and heightened investor interest increased both the probability and pace of deals already in motion, which is what drove the stronger 2027 projection, he said. To be sure, usage of the most advanced closed source AI models has been unprecedented this year, and faster improvements to such models could peel away open source model users. Sales of Anthropics' closed source models have increased more than five times in the first half of the year to roughly $4 billion per month, a pace of $50 billion per year. Some of the increase came from large customers such as Meta Microsoft and Amazon. But open source models,

especially those from China, are undoubtedly improving developers say even if some of them struggle with follow-up questions or deeper lines of reasoning. Data from OpenRouter, a service that compares AI models so developers know which ones should power their apps or handle different tasks within them show that the share of all tokens processed on open source models, primarily DeepSeek and Minimax from China, climbed to roughly 65% in the second week of June, up from 34% during the same week in January and 26% the year prior, end quote. Scotland plays Brazil tonight, and Scotland needs a result or they are going home.

Let me tell you my favorite story about Scotland playing Brazil. The two countries faced each other in 1982, and before the game, referee said, good luck, may the best team win to both of the team captains, to which Scotland Captain Graham Soonis replied, I bloody well hope not. Talk to you tomorrow.