Tech Brew Ride Home - Tue 08/05 – Spygate At TSMC?

Episode Date: August 5, 2025

TSMC has fired employees for allegedly attempting to pilfer information on their 2nm tech. Figma could have raised more money in its IPO but chose not to. Creating entire video game worlds with just a text prompt. And exactly how big Patreon has grown in the creator economy. 00:32 TSMC Spygate 02:50 Cloudflare V. Perplexity 04:52 Figma Followup 07:35 Music AI 09:53 New Google Models 14:58 Patreon Numbers Links: TSMC says employees tried to steal trade secrets on iPhone 18 chip process (9to5Mac) Cloudflare says Perplexity’s AI bots are ‘stealth crawling’ blocked sites (The Verge) Figma’s Pursuit of Long-Term Backers Kept IPO Price in Check (Bloomberg) Voice Startup ElevenLabs Launches AI Music Service (WSJ) Google’s new AI model creates video game worlds in real time (The Verge) Google’s Kaggle to host AI chess tournament to evaluate leading AI models’ reasoning skills (Silicon Angle) Exclusive: Patreon crosses $10 billion creator payment milestone (Axios) Learn more about your ad choices. Visit megaphone.fm/adchoices

No one goes to Hank's for his spreadsheets. They go for a darn good pizza. Lately, though, the shop's been quiet. So Hank decides to bring back the $1 slice. He asks Copilot in Microsoft Excel to look at his sales and costs to help him see if he can afford it. Co-pilot shows Hank where the money's going and which little extras make the dollar slice work.

Now, Hank has a line out the door. Hank makes the pizza. Co-Pilot handles the spreadsheets. Learn more at M365Copilot.com slash work. Welcome to the Tech Brew Ride Home for Tuesday, August 5th, 2025. I'm Brian McCalla today. TSM has fired employees for allegedly attempting to pill for information on their two nanometer tech. Figma could have raised more money in its IPO but chose not to, creating entire video game worlds with just a text prompt, and exactly how big Patreon has grown in the creator economy.

Here's what you missed today in the world of tech. Weirdly, I have two sort of SpyGate stories for you today. First up, Niki Asia says, TSM has fired several employees for violating rules on obtaining sensitive information about its two-nanometer technology. TSM says it, quote, detected unauthorized activities. Quoting 9 to 5 Mac. TSM leads the world in the most advanced chip processes and is next year expected to use its two-nanometer technology for the A20 chips used in the iPhone 18 range. Apple typically gets access to TSM's most advanced.

chip processes ahead of the company's other customers. Apple analyst Ming Chi Quo has suggested that the new chip will be used for all iPhone 18 models, not just the two pro ones. Nikai Asia reports that TSM accused several former employees of attempting to obtain secret information about its two nanometer chip development and production process. Several former employees of TSM are suspected of attempting to obtain critical proprietary information on two nanometer chip development and production while working at the company, according to multiple sources familiar with the matter. In response to Niki Asia's questions, TSM said that it recently detected unauthorized activities during routine monitoring, leading to the discovery of potential trade secret leaks.

The world's top chipmaker said on Monday it took strict disciplinary actions against the personnel involved and has initiated legal proceedings. The attempt was detected by spotting unusual access patterns on the part of one of the employees. The report says there could even be national. security implications as the Taiwanese government takes extremely seriously the protection of advanced technology developed within the country. Prosecutors have confirmed that they are investigating, and TSM says that it will seek prosecution to the fullest extent of the law. No details have been shared on the nature of the information obtained. It is likely that it relates

to the two nanometer process in general rather than anything specific to Apple's A20 chip, end quote. late news that Taiwanese prosecutors say two XTSMC employees and a third suspect have now been arrested over this, and investigators reportedly searched the offices of Tokyo Electron, a Japanese chip toolmaker. And then remember how Cloudflare has been leading a charge against AI bot scraping? Well, Cloudflare now says perplexity uses stealth crawling techniques like undeclared user agents and rotating IP. addresses to evade robots. Text files and network blocks, quoting the verge. The report only adds to concerns about Perplexity vacuuming up content without permission as the company got caught barging past paywalls and ignoring sites robots.comptex files last

year. At the time, Perplexity CEO Arvin Srinivas, blamed the activity on third-party crawlers used by the site. Now, Cloudflare, one of the world's biggest internet architecture providers said it received complaints from customers who claim that perplexity's bots still had access to their websites even after putting their preference in their websites, robots.text file, and by creating web application firewall rules to restrict access to the startup's AI bots. To test this, Cloudflare says it created new domains with similar restrictions against perplexity's AI scrapers. It found that the startup will first attempt to access the sites by identifying itself as the names of its crawler's perplexity bot or a perplexity user. But if the website has

restrictions against AI scraping, Cloudflare claims perplexity will change its user agent, the bit of information that tells the website what kind of browser and device you're using, or if the visitor is a bot, to, quote, impersonate Google Chrome on MacOS. Cloudflare says this undeclared crawler uses rotating IP addresses that the company doesn't include on the list of IP addresses used by its bots. Additionally, Cloudflare claims that perplexity changes its autonomous systems network, a number used to identify groups of IP networks controlled by a single operator to get around blocks as well.

This activity was observed across tens of thousands of domains and millions of requests per day, Cloudflare writes, end quote. Just wanted to do a quick follow-up on the whole Sigma IPO thing. Figma's stock did drop 27.3.3.5.5.5.5.5. to $88.60. Yesterday, probably an inevitable bounce after last week's IPO pop. But its fully diluted valuation still sits around $56 billion, so still almost triple Adobe's 2022 acquisition offer. And on that hole, did they leave money on the table argument? A source is telling Bloomberg that Figma did consider a higher IPO price, but CEO Dillon Field wanted to bring on board certain long-term

institutional shareholders and approved the $33 per share pricing at the IPO open. The stock's 250% jump on its debut session, the biggest ever first day pop from a billion-dollar-plus US IPO, meant the company and the selling shareholders effectively handed over more than $3.5 billion of value to investors that were able to get shares. Figma and the banks considered pricing the shares above the ultimate $33 per share set in the IPO. according to a person familiar with the matter, but in the end, Dylan Field, Figma's co-founder and chief executive officer, wanted to bring on board certain long-term institutional shareholders and Field signed off on the final $33 price.

The person said asking not to be identified as the information isn't public. With shares oversubscribed by more than 40 times, a blockbuster debut was by no means unpredictable. Had Figma and Field opted for a higher price, however, some institutional investors may not have been willing to buy the shares. firms with a long-term investment strategy often base their view of the stock on fundamentals and determine their price expectations in reference to the valuations of comparable software firms. These investors are typically the most coveted backers in a first-time share sale, as they tend to bring price stability.

The bank's pitched Figma's IPO valuation in line with highly valued software companies such as Snowflake and Data Dog that trade at forward enterprise value to sales multiples ranging from 10 to 15 times, according to data compiled by Bloomberg. Silicon Valley venture capitalist Bill Gurley, general partner of benchmark, was among the most vocal critics of the pricing. Figma's day one pop highlighted the, quote, gross inefficiency in the modern IPO process. He wrote Thursday in a post on X, a frequent commentator on the phenomenon. Gurley has long advocated for companies to pursue direct listings, raising no money and letting the market determine the share price to avoid leaving large sums on the table.

Gurley wasn't listed as a shareholder in Figma's filings, end quote. Long time, listeners will remember 11 labs as that AI startup that I first experimented with to clone my voice years ago now. Well, if I ever need new music for the pot, I guess I'll go back to them because they have launched 11 music, letting individuals and businesses use their AI model to generate music cleared for commercial use via English language prompting. Quoting the journal, users enter a prompt. in plain English, such as create a smooth jazz song with a 60s vibe and powerful lyrics, but

relaxing for a Friday afternoon, and the startup's AI model generates a tune within minutes, complete with vocals and instrumentals. With the launch, 11 Labs, best known for its voice generation software, enters a frat sphere where major music labels have already sued two music generation startups, Suno and UDO, for their alleged use of copyrighted works to train their AI. 11 Labs co-founder and chief executive Maddie Stanisleski said the three-year-old startup has a deal with Merlin Network, a digital rights agency for independent labels, to train its models on artists' work, whose rights are represented by Merlin. Eleven Labs has a similar deal with Cobalt Music Group, an independent rights management and music publishing firm. Both Merlin and Cobalt are still finalizing which artist's music will be involved in the training for Eleven Labs' music AI model. The company said, those deals give the models.

AI created music legal cover for broad commercial use. Stanisluski said, he hopes to also strike deals with the major labels, including Universal Music Group, Sony Music Group, and Warner Music Group, he said. Currently, we aren't using their data in our model, Stanisluski said, referring to the major labels' music trove. The model is strictly created on data we have access to. Eleven Labs has also built safeguards, it said, which prevent its model from creating songs with artist's names

or specific lyrics from an album or label. it also blocks lyrics that could incite violence or are obscene or unlawful. 11 music is geared toward both businesses and creators, the company added, allowing them to produce their own music for commercial projects or creative endeavors without the need to hire a composer or pay for a restricted use license from a major label, end quote. Week pollination season, and my business is scaling fast. To keep the nectar flowing, I need a phone plan with top priority data speeds. That's why I chose GoogleFi Wireless.

My connections stay strong, even when the high. is buzzing. Plus, unlimited plans started $35 a month. Now that's a deal that doesn't stay. Explore GoogleFi Wireless plans today. Plus taxes and government fees. GoogleFi Wireless is not subject to data traffic deprioritization during times of high network usage. Ready to soundtrack your summer? With Red Bull Summer All Day Play, you choose a playlist that fits your summer vibe the best. Are you a festival fanatic, a deep end DJ, a road dog, or a trail mixer? Just add a song to your playlist and put your summer on track. Red Bull Summer All Day Play.

Red Bull gives you wings. Visit redbull.com slash bright summer ahead to learn more. See you this summer. And Google is still throwing new AI stuff out there. Google Deep Mind has released its Genie 3 model, which can generate 3D worlds from a prompt and has enough visual memory for a few minutes of continuous interaction. Quoting the verge,

world models are a type of AI system that can simulate, for purposes like education, entertainment, or to help train robots or AI agents. With world models, you give them a prompt and they generate a space that you can move around in like you would in a video game. But instead of the world being handcrafted with 3D assets, it's all being generated with AI. It's an area Google is putting a lot of effort into the company showed off its genie 2 in December, which could create interactive worlds based off an image. And it's building a world models team led by a former co-lead of OpenAI's SORA Video Generation tool. But the models currently have a lot of drawbacks.

Jeannie Two Worlds were only playable up to a minute, for example. I recently tried interactive video from a company backed by Pixar's co-founder, and it felt like walking through a blurry version of Google Street View, where things morphed and changed in ways I didn't expect as I looked around. Jeannie 3 seems like it could be a notable step forward. Users will be able to generate worlds with a prompt that supports a few minutes of continuous interaction, which is up from the 10 to 20 seconds of interaction possible with Jeannie. 2, according to a blog post. Google says that Genie 3 can keep spaces in visual memory for about a

minute, meaning that if you turn away from something in a world and then turn back to it, things like paint on a wall or writing on a chalkboard will be in the same place. The worlds will also have a 720p resolution and run at 24 frames per second. DeepMind is adding what it calls promptable world events into Genie 3, too. Using a prompt, you'll be able to do things like change weather conditions in a world or add new characters. However, this probably isn't a model you'll be able to try for yourself. It's launching as a limited research preview that will be available to a small cohort of academics and creators so its developers can better understand the risks and how to appropriately mitigate

them, according to Google. There are also plenty of restrictions like the limited ways users can interact with generated worlds, and that legible text is often only generated when provided in the input world description. Google says it's exploring how to bring Jeannie 3 to additional testers down the line, end quote. And on top of that, Deep Mind has also unveiled what it is calling Kaggle Game Arena, a benchmarking platform where AI models will compete head-to-head in strategic games starting with a chess tournament this week. Quoting Silicon Angle, the three-day AI chess battle is the first in a series of tournaments

set to be hosted by Google's data science community Kaggle within a newly developed game arena. There, the models will compete against each other in a range of strategic games designed to evaluate their thinking and reasoning capabilities. Google DeepMind and Kaggle are partnering with Chess.com, the chess app Take, Take, and legendary chess live streamers Levi-Rosman and Hikaru Nakamura on the tournament with the first simulation set to begin tomorrow. The Kaggle Game Arena is a new AI benchmarking platform that's designed to test how competitive large language models are in a series of strategic games, including Go and Werewolf. But first up is the AI Chess Exhibition, which runs August 5th through 7th with the simulated games live streamed on caggle.com.

Hikaru Nakamura will provide commentary on each of the matchups, while Levi Rosman will deliver a daily recap of each day's battles, complete with analysis on the Gotham Chess YouTube channel. The tournament will conclude with a stream of the championship matchup and tournament recap from Magnus Carlson on the Take, Take, YouTube channel. There will be eight competitors battling for chess supremacy, Gemini 2.5 Pro, Gemini 2.5 Flash, Claude Opus 4, Deepseek R1, Moonshots, Kimi 2K2 Instruct, 03, 04 Mini, and Grock 4. The tournament will be based on a standard single elimination bracket format, where the winners of each match will be decided over a Best of Four series of games. Kaggle Game Arena will live stream one round each day, so the first round will involve four matchups of eight models at the quarter final stage, followed by two matchups in the semifinal

round on day two and a single final matchup on day three. In a blog post, Google outlined a number of rules saying that the models will be responding to text-based inputs. None of the competing models will be allowed to access any third-party tools so they can't just use the stockfish chess engine to identify the best move in any situation. Instead, they'll have to think about it themselves. The models will not be given a list of possible legal moves, and if one attempts to make such a move, it will be allowed three retries. Should it fail to make a legal move, it will forfeit the game. Moreover, there will be a 60-minute time limit for each move. The livestream will attempt to show how each of the competing models reasons about its next move

and its response to any failed moves, end quote. Finally today, Patreon CEO Jack Conte says the platform has paid out now more than $10 billion to creators since its 2013 founding, and creators now give. it more than $2 billion annually, and there are now more than $25 million paid memberships. Quoting Axios, when Patreon launched in 2013, it focused mainly on facilitating payments between creator-fan relationships that were mostly built on other platforms. Today, it offers new tools and products to help creators build and grow their audiences on Patreon. Patreon's iOS app is now the company's top platform for fan engagement, a spokesperson said. This year, Patreon released a slew of

new podcast tools, including multi-show management, RSS mirroring, and syncing. It's currently testing a new feature to allow creators to upload ad-free live video to the platform directly. What we're seeing in creators is that just like media companies, they're multi-hyphenate, Conti says. That multimedia multi-hyphenate approach works really well on Patreon's infrastructure because we've spent 10 years building horizontally across categories. Last year, Patreon creators were paid $472 million on podcasts alone, and fans' watch. watched over 80 million hours of video on the platform. For years, Patreon was focused solely on paid subscriptions. In 2023, it expanded to allow creators to offer free memberships to fans to broaden

their audience of potential subscribers. That shift combined with its product investments has helped grow the number of paid subscribers on the platform significantly, Conti says. Today, there are over 100 million free memberships on the platform per Conte, with more than 700,000 members converting from free tiers to paid per month. Patreon is still private. held and Conti says he has no plans right now to take the company public. The firm, which has raised more than $400 million to date, was most recently valued at over $4 billion in 2021, end quote. Hey, y'all, the official relaunch as Tech Brew Ride Home has gone well so far as I write this. We are number two in the technology category on Apple Podcasts just behind Acquired.

But ahead of All In and Lex Friedman, I pinged David and Ben this morning, and I said, I'm coming for them. I'm coming for you. Acquired, help me out. Let's get to number one. One way you can do that, since you're already subscribed and listening to this, is to leave us a review. If you've never left a review for this podcast before, stop right now, go into your podcast app and leave us a five-star review and a few words about us. If we bombard the pod rankings with fresh reviews, that helps boost our place. So give it a go. Give us a review if you've never done so before. Thanks in advance. Talk to you tomorrow. Yamava Resort and Casino at San Manuel is California's number one entertainment destination

for today's superstars. Catch the Jonas Brothers return to the Yamava Theater stage on April 30th, the powerful vocals of Demi Levato on May 17th, and the signature Southern Country Rock of Eric Church on July 19th. Tickets on sale now at Yamavatheater.com, only at Yamava Resort and Casino, celebrating its 40th anniversary. You in? Must be 21 to enter.