Tech Brew Ride Home - The New GPT

Starting point is 00:00:00 Here we go. Someone's already claiming this is our year. Someone else said that last year too. A round of Jameson, Ginger, and Lime arrives at a table. Smooth enough for kickoff, smooth enough for extra time. New friends pulling up a stool. Debates about whether that was a handball. Cheers rising like a roar around the room.

Starting point is 00:00:20 Because match days are about the shared moments. How did Jameson to your match day lineup? Jameson, it's what you bring. Please enjoy our products responsibly. Welcome to the Tech Roo Ride Home for Friday, July 10th, 2026. I'm Brian McCullough today. OpenAI released GPT 5.6 and launched Chat GPT work. Vigicc stepped down from OpenAI, citing health reasons. The EU found meta's addictive design violates the DSA. Polymarket sought margin trading approval and SKHinnix debuted on the NASDAQ. Here's what you miss today in the world of tech. Every day shareholders meet to discuss important matters about the companies you invest in. Now you can make your voice heard too. Vanguard Investor Choice makes it easy to set your proxy voting preference for your eligible Vanguard index funds,

Starting point is 00:01:16 whether you hold a Vanguard fund directly or through another brokerage firm. All it takes is a few clicks to select your proxy voting preference and be heard on important shareholder topics like executive pay and director elections. Visit vanguard.com slash investor choice to learn more. It's your shares. It's your voice. It's easy. Vanguard investors own shares in our index funds, and those funds own shares of the companies they invest in, Vanguard Marketing Corporation distributor. OpenAI released GPD 5.6 broadly yesterday and launched ChatGPT Work, an AI agent that can gather content across apps and files to create documents on MacOS and Windows.

Starting point is 00:01:58 Quoting TechCrunch. GPD 5.6 comes in three variants. Seoul, considered its workhorse, Terra, a more interesting. intermediate option, and Luna, its budget-friendly option. These models expand what users can do across a variety of fields, with the company promising powerful capabilities and enterprise work, coding, and even scientific research. CEO Sam Altman has promised that his company's newest models are orders of magnitude more efficient and cost-effective than previous versions, recently telling CNBC that Seoul is 54% more token-efficient when it comes to AI coding tasks.

Starting point is 00:02:28 Most notably, the company calls 5.6 its strongest cybersecurity model yet, achieving frontier performance with significantly fewer tokens. Indeed, much hubbub has been made about the model's cyber capabilities, as the Trump administration previously sought to restrict its rollout, ostensibly due to fears of how the model could be misused. GPD 5.6 supports defensive activities, including threat modeling, code review, and patching, and blue teaming, simulating an attack on your own systems to find weaknesses before real hackers do. OpenAI also released a new tool called ChatGPT Work, which, just as it sounds, is designed as a workplace companion for enterprise teams running on desktop, web, and mobile that can help with daily clerical tasks like drafting documents,

Starting point is 00:03:11 spreadsheets, and presentations. GPD 5.6 and its attendant marketing seems most designed to take aim at OpenAI's primary opponent Anthropic. Anthropic has managed to make itself the likable underdog of the AI race focusing fixedly on enterprise customers and winning a growing share of support as a result. Not to be outdone, OpenEy I cites the artificial analysis coding agent index, a notable benchmarking metric to claim that its latest family of models outshines Anthropics models at every turn. Open AI calls Sol its best coding model yet and has explicitly compared it to Anthropics recently released and much-hyped Fable using the coding agent index. Open AI claims that Sol sets a new state of the art at 80, 2.8

Starting point is 00:03:54 points above Fable 5, while using less than half the output tokens, taking less than half the time, and costing about one-third less. That advantage extends across the family. Terra performs just above Fable 5, while Luna outperforms Opus 4.8. The company says that 5.6 is now available across ChatGBT, GBT, Codex, and the OpenAAI API. Availability per million tokens is priced as follows. Seoul is $5 per input, $30 per output. Terra is $250 per input, $15 per output, and Luna is $1 per input, $6 per output, end quote. Fiji Simo, OpenAI's CEO of AGI deployment, says she will step down and become a part-time advisor to the company after her medical condition has apparently worsened.

Starting point is 00:04:48 Simo joined OpenAI in August, quoting the journal. She communicated her decision in a note to staff Thursday saying that her medical condition had worsened and her road to recovery would be much longer than anticipated. She will become a part-time advisor to the company. Simo joined last August and took on many managerial responsibilities from Sam Altman, including overseeing the chief financial officer and chief revenue officer. She said in April that she was taking medical leave, telling staff that her neuroimmune condition had worsened. Her departure created a leadership vacuum at the time, raising questions among investors and employees about the future direction of the company. CMO is tapped to accelerate chat GPT's growth, and she made her mark by introducing ads to the popular chatbot, as well as features such as health advice.

Starting point is 00:05:33 She had previously served as the CEO of Grocery Delivery App Instacart and, as a top executive at Facebook, making her a strong choice to oversee the growth of OpenAI's consumer-facing business. But ChatGPT's growth stalled toward the end of last year, contributing to missed internal revenue goals. The company abruptly pivoted its focus to building AI-powered coding tools for businesses after falling behind Anthropic in that lucrative market. Seymot led early efforts to create a coding-focused super app, which OpenAI launched yesterday and cut side projects such as the video generator app SORA. At the Allen and Company retreat in Sun Valley, Idaho this week, Open AI executives discussed how the company plans to compete with Anthropic to win over more enterprise customers, according to a person who spoke with them.

Starting point is 00:06:18 In her internal memo Thursday, Simo said that her product and business responsibilities would be divided among President Greg Brockman, CFO, Sarah Fryer, and Chief Strategy Officer Jason Kwan. Denise Dressor, OpenAI's chief revenue officer, will now report to Brockman. This has been one of the hardest decisions of my career, but my body left me no choice, my seat. symptoms became as loud as I am stubborn, she wrote. Simo said in a post on X Thursday that when she went on leave, people told her that she was courageous for prioritizing her health. The truth is that I am only making this decision now because I failed to make it many times before, she wrote, end quote.

Starting point is 00:07:00 In preliminary findings, the European Union Commission said Facebook and Instagram's quote, addictive design violates the DSA. telling META to make changes or risk hefty fines. Quoting the Times. Regulators in Brussels said the addictive design of META's two social media services violated European Union law and that it needed to remove features like infinite scroll and the automatic playing of videos. The technology giant should also introduce new, quote, screen time breaks and adjust its

Starting point is 00:07:32 recommendation algorithm to make it, quote, less engagement-oriented, authorities said. The Commission's findings are preliminary and META now has time to respond to the allegations before a final judgment is issued. A fine could run up to 6% of META's global revenue, though regulators rarely issue penalties of that size. The ruling is an unusually direct effort to force specific product design changes. It is the latest sign of the aggressive approach taken by European regulators to confront social media companies over what is seen as intentionally subversive efforts to hook users, especially children. In February EU authorities accused TikTok of using similar design tricks to keep users coming.

Starting point is 00:08:10 back again and again. Ben Walters, a spokesman for META, said the company disagreed with the findings, which, quote, don't accurately take into account the significant steps we've taken to protect teens. He said the company has introduced teen accounts that allow parents to block access to Instagram at night and cap daily screen time. The European Commission, the executive branch of the 27-Nation block, is also weighing rules that could bar children from social media entirely because of the app's addictive features following Australia's lead. In Brussels, officials have said they are weighing where to draw the line between a well-designed app that users enjoy and a harmfully addictive one. The European Union, home to roughly 450 million people, is one of the world's largest

Starting point is 00:08:49 markets for social media, but authorities there have become increasingly skeptical of the ways the companies operate. Authorities said features like personalized recommendations and an infinite scroll that constantly deliver new content fuel the user's urge to keep scrolling and shift the brain into autopilot mode, their words. Such features lead to, quote, unhealthy habits and compulsive use, the authorities added. Regulators said meta did not adequately assess the risks of its products, including ignoring information about how much time users under the age of 18 spent on Instagram and Facebook at night. At the same time, features like reels and stories led to, quote, excessive or compulsive use of the services, authorities said. Meta's existing time

Starting point is 00:09:29 management tools are easily bypassed by younger users and its parental controls are effective only if parents have technical savvy and can devote considerable time to monitoring regulators said. Protecting the physical and mental health of Europeans must be a priority for social media platforms. Hannah Vakunin, the executive vice president at the European Commission overseeing digital policy said in a statement, the Digital Services Act provides a clear framework to hold platforms accountable for the addictive design and effects of their services. We are fully committed to enforcing our legislation in Europe, end quote. Ever spent hours of your workday tracking down information only to find that it wasn't documented at all?

Starting point is 00:10:14 It's an age-old corporate conundrum. Critical knowledge isn't available or accessible to the folks who need it. That's the very problem our sponsor, Scribe, was built to fix. Scribe is a workflow AI platform that automatically turns your workflows into clear-cut documentation. All you do is turn on the Scribe browser extension or desktop app and go through your processes as you would normally. Scribe builds a guide as you go, capturing every click, step, and screenshot automatically. So what would have taken hours gets done in under a minute. To see what Scribe could look like for your org, head to Scribe.

Starting point is 00:10:48 Dot how slash ride home. And mention ride home for your first month of Scribe Capture Free on Select Plans. That's S-C-R-I-B-E dot how slash ride home. When your company deploys a customer-facing AI that misses its mark, who do you think those customers blame? Well, according to the 2026, Delight AI Index, 83% of customers survey blamed the brand, not the tech they were using. That's why it's important to have tech you can trust. Delight.A.I. is a customer-facing AI that delivers hyper-personalized experiences on your behalf.

Starting point is 00:11:26 With zero-touch improvement, it can continuously monitor its own performance, find failure patterns before they spread, write the fix, and ship it. It gets smarter with every conversation automatically. The longer it runs, the more your customers can trust it. To learn more, just head to Delight.A.I. slash brew. That's Delight.com.A.I. slash brew. Why do I have the increasing feeling that this will all end badly someday? Polymarket is seeking CFTC approval to offer margin trading in the U.S. A move that would let users bet on events on the platform with less capital up front. Quoting Bloomberg, Pollymarket must obtain approval from the commodity.

Starting point is 00:12:10 Futures Trading Commission for changes to its rulebook that would allow non-fully collateralized trading. Institutional investors see margin trading as a more capital-efficient way to place bets across markets. CalShe got its own FCM license earlier this year in a bid to expand client access to event bets on its platform at a larger scale, and it is part of the infrastructure needed to support perpetual futures. Perpetual futures are a type of derivative popular in crypto markets that, unlike traditional futures contracts do not expire. Instead, the value of the contract remains tied to the underlying asset, creating a tradable instrument that tracks the current price of a commodity. Kalshi racked up more than $5.5 billion worth of perpetual's trading volume within two weeks

Starting point is 00:12:53 of the product's official launch. The company offers contracts primarily tied to crypto tokens. Retail traders have turned prediction markets into one of the fastest growing parts of finance, with weekly notional volume on polymarket hitting a record of more than $4 billion in June. The platforms are also attracting investment from traditional exchange operators and getting institutional investors on board is seen as a crucial next step in their development. Polymarket uses a public blockchain ledger that makes it possible for anyone to see individual bets, though the associated accounts are generally suit anonymous. While suspicious trades have long surfaced across financial markets, the traceability of unusually well-timed trades on Polymarket has drawn scrutiny from lawmakers and regulators over insider trading risks in the prediction market industry. U.S. regulators would require users accessing margin products on any prediction market platform to undergo additional identity checks, including providing employer information.

Starting point is 00:13:45 Those requirements have become a hot-button issue as the platforms grapple with high-profile instances of insider trading, end quote. When companies in foreign countries list their stock in the U.S., it's kind of like an IPO, but just for this country. S.K. Heenix has done just that, raising $26.5 billion in the U.S. the largest ever U.S. market debut by a foreign company, selling 177.9 million ADRs for $149 each. Those are American depository receipts, essentially IOUs on foreign stock. The sale was more than seven times oversubscribed, and it opened up at 14 percent at $170 per share,

Starting point is 00:14:34 quoting CNBC. It's kind of a dream, and now it's a dream come true. S.K. Heenix Chairman Che, Che, Taiwan told CNBC's, Christina Parts of Nevelis on Friday. SKHenix trails only Samsung by market cap in its home country. Like its larger rival, the company makes computer memory, which is used by phones and PCs to store short-term data. SKHenix's roster of customers include some of the biggest names in technology, such as Nvidia and Apple.

Starting point is 00:15:01 Memory for decades was tucked in a sleepy corner of the semiconductor world, but the artificial intelligence boom has turned it into a massive growth market. Taiwan told CNBC that when he meets with it, customers and partners, everybody expects more chips. He said that even when SKHenix announced it would double capacity within five years, customers said they still need more. All my customers said that, well, that's not enough, man, and well, we need more. Taiwan said Friday. SKHenix's valuation has risen more than sevenfold over the past year, as demand for AI infrastructure has caused a shortage in computer memory and sent prices skyrocketing. SKHenix is the leader in the high-performance

Starting point is 00:15:38 its memory that's used in AI chips from Nvidia, the world's most valuable company. Compared to RAM for phones or laptops, AI chips require high bandwidth memory or HBM, which is created through a complex process that involves stacking many layers of traditional memory together. Some of that HBM will be packaged in the U.S. after the company announced a $4 billion advanced packaging plant in Indiana. But the vast majority of S.K. Hennox's planned expansion over the coming years will take place in South Korea. That includes a cluster of chip fabrication plants. that will cost around $390 billion, end quote. Finally today, another new smart glasses entrant,

Starting point is 00:16:21 and this one is making me get an itchy trigger finger hovering over the buy button, quoting the verge. I love it when a company challenges itself to make a cheaper version of a beloved product. X-Reel's $299 A-1 Plus is a stripped-down version of its $449 1S. That's light on features, but with just enough of the 1S,000. best qualities. These AR glasses are comfortable, they look good, and the screens are surprisingly bright and contrast-rich for the price. The new glasses are super lightweight at 62 grams, over 20 grams lighter than the 1S. They're as light as I wish every model was, but the frame feels flimsy

Starting point is 00:17:02 by comparison. Its durability came into question when I adjusted the A01 Plus's temple arms to center the screens. They require more of a gentle touch so as not to over-muscle the frame that houses its delicate micro-OLED panels and bird bath optics or pop off its modular eyeglasses shell. Speaking of those shells, customization is a big selling point for the A01 Plus. X-Real includes a shell that puts tinted lenses on the other side of its screens, and it can be swapped with other covers, even 3D-printed ones in the future. Removing them requires gently tugging on its edges near where the temple arms attach, which feels risky to do until you get the hang of it.

Starting point is 00:17:40 The company provides an alternate cover that accentuates the A1 plus's bright screens and keeps reflections out of view. The new AR glasses lack electrochromatic lens tech with adjustable opacity, which is standard across most models, but I don't mind the omission here. The AO1 plus have reflective films covering the back of their optics that are passable at preventing light from disrupting the experience. One of the removable covers included with the review unit is better at blocking out light than the Pricer 1S. I held my phone's flashlight right up to the lens as I wore them, and I could barely see it. As with every pair of AR glasses that I've tested, I played a lot of games from my

Starting point is 00:18:17 Steam deck on the A1-1-plus. First impressions with these glasses were mixed. The 1080 pre-resolution, brightness, 120-hertz refresh rate, and contrast looked good, but the screens appeared very blurry. However, X-R provided a set of Hans VR prescription lenses that drastically improved clarity. The glasses have a IPD range of 54.5 millimeters to 74.5 millimeters, but I can't. guarantee that you'll have a better experience than I did out of the box unless you buy some lens inserts, which will likely cost around 50 bucks. Despite some shortcomings, the A-O-1-plus glasses are exactly what some people will be after, a relatively affordable set of USBC AR glasses for mirroring their games, movies, or whole computer screens on. They're lighter on features,

Starting point is 00:19:02 but they nailed the basics with bright screens, serviceable audio quality, and a comfy, lightweight build. I am actually impressed with all that's included for the price. The features, you get by spending $150 more for the 1S might be worth the plunge for some, but generally the A01 plus are great for first-time buyers who don't want to spend more than $300. Just make sure to factor in the $50 you may need to spend on prescription lenses, end quote. So number one, I just realized I forgot to do a long read section. I blame jet lag and me not really knowing what day it is, so sorry about that. And number two, does anyone listening know anyone at?

Starting point is 00:19:47 Anthropic. I want to get in touch with them about maybe doing an internet history podcast episode with them. If anyone can make an intro to Anthropic, please hit me up at Brian at ridehomefund.com. Thanks in advance. Talk to you on Monday.

Tech Brew Ride Home - The New GPT

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.