← Tech Brew Ride Home
Tech Brew Ride Home
Does Anthropic Owe Me $3000?
00:24:26
Anthropic has agreed to pay $1.5 billion dollars to authors of books that might have trained their AI. Would my book qualify for some of that money? Also, why this is interesting in general. OpenAI is making a movie. Tokenizing the stock market. Three interesting raises, and at the end of the show, let me tell you about my weekend experiment with AI.
Loading summary
Transcript11 lines
- [00:04]
A
Welcome to the Tech Brew. Write home from Monday, September 8, 2025 I'm Brian McCullough. Today, Anthropic has agreed to pay one and a half billion dollars to authors of books that might have trained their AI. Would my book qualify for some of that money? Also, why this is an interesting thing in general. Also, OpenAI is making a movie tokenizing the stock market, three interesting raises, and at the end of the show, let me tell you about my weekend experiment with AI. Here's what you missed today in the world of well, this would be one of the reasons they raised all that money recently, I guess. Anthropic has agreed to pay one and a half billion dollars plus interest to resolve an author's copyright lawsuit over the company's downloading of millions of pirated books. Quoting Bloomberg A request for preliminary approval of the accord involving one of the fastest growing AI startups was filed Friday with a San Francisco federal judge who had set the closely watched case for trial in December. The settlement is among the in dozens of copyright lawsuits filed against AI leaders, including OpenAI, Meta Platforms and Midjourney, and alleging misuse of proprietary online content. Anthropic had said in a court filing that it felt inordinate pressure to cut a deal to avoid a potentially business ending trial that could have put the company on the hook for as much as $1 trillion in damages. The startup recently reached $5 billion in terms of run rate revenue and raised $13 billion in investment at a $183 billion valuation. But Anthropic is still ultimately unprofitable due to the high costs of developing AI, a lawyer representing the plaintiffs said. The accord far surpasses any other known copyright recovery. This settlement sends a powerful message to AI companies and creators alike that taking copyrighted works from these pirate websites is wrong, justin Nelson of Sussman Godfrey said in a statement. The case was brought as a class action on behalf of the authors of as many as 7 million books, who claimed that the startup illegally used downloaded, pirated versions of their copyrighted texts to train its large language AI models, even though it was unclear whether those materials were actually used for training. A lawsuit of that size could have driven the company to bankruptcy if it were to lose in trial. Under the terms of the deal, Anthropic will pay about $3,000 for each of about 500,000 books in the class. If more claims are submitted, the total payout would grow. The company also agreed to destroy data it was accused of illegally downloading. A hearing on the proposed settlement is set for September 8th. Hmm. I wonder if I should apply for this. $3,000 ain't chump change, and I'm sure my book on the history of the Internet is in there. I did an AI project this weekend and it cited research that I did years ago. More on that at the end of the episode. But MG Siegler makes a great point about this case specifically. Quote you can't help but wonder if part of the equation in the settlement wasn't decidedly cynical. Fresh off a new massive fundraise, one in which they raised far more than they were initially targeting, Anthropic has a lot of money. More than perhaps all but one of their competitors on the startup side. By settling for $1.5 billion, is anthropic sort of pulling up a drawbridge, making it so that other startups can't possibly come into their castle? I mean, am I crazy? I'm not sure I am. At $1.5 billion, there are only a handful of companies that could afford to pay such. Certainly OpenAI is one, maybe Xai, and of course all the tech giants like Apple, Amazon, Microsoft, Google and Meta. But could any other startup that has done any level of model training with such data? Probably not. And so I wonder, did Dario Amodai just pull a Michael Corleone like elimination of enemies move overnight? Was this a maneuver he learned from elon musk? Oh, OpenAI wants to convert into a for profit. Let's set a new floor price on that. Dario just set a floor price on AI training infringement. This is a message to all competitors. A standard has been set. Open up those wallets if you want to keep playing this game. I'm watching a lot of journalists and writers seem almost euphoric at the news. How much is my work worth Again? Not so fast. This move may actually jolt the market in a bad way by taking out would be competitors for your words. Anthropic says it's settling because they didn't want to risk having to spend trillions. That may or may not have come to pass. But certainly Anthropic was more likely to have to spend trillions if the AI wars continue on as they have been these past couple years, battling so many companies on so many fronts with this move, Anthropic may have just narrowed that field. End quote. Meanwhile, full speed ahead, I guess in terms of creation in another direction, OpenAI is providing its tools and computing resources for Critters, a largely AI made feature length animated film that is set to hit theaters globally in 2026. Quoting the journal Critters, about forest creatures who go on an adventure after their village is disrupted by a stranger, is the brainchild of Chad Nelson, a creative specialist at OpenAI. Nelson started sketching out the characters three years ago while trying to make a short film with what was then OpenAI's new Dall E image generation tool. Now he has teamed up with production companies in London and Lo Angeles, aiming to debut a feature length version of the film at the Cannes Film Festival in May. The team is attempting to make the movie in about nine months instead of the three years it would typically take, said James Richardson, co founder of London based Vertigo Films. Vertigo is producing the film along with Native Foreign, a studio that specializes in using AI along with traditional video production tools. Critters has a budget of less than $30 million, far less than what animated films typically cost. The production team plans to cast human actors for character voices and higher art to draw sketches that are fed into OpenAI's tools, including GPT5 and image generating models. OpenAI can say what its tools do all day long, but it's much more impactful if someone does it, nelson said. That's a much better case study than me building a demo. The script for Critters was written by some members of the team that wrote Paddington in Peru. Production has begun and casting decisions for the voices of the characters will be made in the next few weeks. It is funded by Vertigo's Paris based parent company Federation Studios. The studios involved are developing a compensation model to the roughly 30 people working on Critters to share in any profits, Nelson said OpenAI is betting that if Critters is successful, it will show that AI can deliver content strong enough for the big screen and accelerate Hollywood's adoption of the technology, Nelson said. OpenAI's tools also lower the cost of entry, allowing more people to make creative content, he said. An OpenAI spokesman said the film reflects the kind of creativity and exploration we love to encourage. End quote. Again, I encourage you to listen to the end of the show because my weekend experiment kind of ties into this. One thing we really haven't gotten a chance to talk about from the crypto world has been the increasing move to tokenize everything. Like you can trade tokenized versions of stock, even stock like OpenAI stock, which is weird because they're not public yet and OpenAI doesn't approve of you doing that. But again, I guess it's full speed ahead. The NASDAQ has reportedly asked the U.S. securities and Exchange Commission to let investors trade tokenized securities on its exchange potentially rolling out the first token settled trades in Q3 of 2026, quoting Reuters. If approved, the move would mark the first instance of tokenized securities being allowed to trade on a major US Stock exchange and also signify the most ambitious attempt yet by an exchange operator to bring blockchain based settlement into the national market system. Nasdaq on Monday filed a proposal with the U.S. securities and Exchange Commission to tweak its rules to allow for trading of listed stocks and exchange traded products on its main market in either traditional digital or tokenized form. The filing comes days after the SEC unveiled its rulemaking agenda, which included a potential amendment of its rules to allow for crypto to be traded on national securities exchanges and alternative trading systems. The latest developments come as investor demand for tokenized assets is rising globally. Proponents of the crypto industry have argued that tokenization can improve liquidity in the financial system. Coinbase, the largest US Crypto exchange, has also previously sought permission from the SEC to offer tokenized equities to its customers. Some major global banks, including bank of America and Citi, have said they could explore launching tokenized assets, including stablecoins. In its filing on Monday, Nasdaq said it believes the market can use tokenization while continuing to provide the benefits and protections of the national market system. Wholesale exemptions from the national market system and related protections are neither necessary to achieve the goal of accommodating tokenization, nor are they in investors best interests, nasdaq said. Some critics of the industry have warned that the frenzy around tokenization could introduce new systemic risks, especially in the absence of stringent regulation. In July, Hester Peirce, a commissioner at the U.S. securities and Exchange Commission who has frequently spoken positively about cryptocurrency, said tokenized securities would not be able to circumvent existing securities laws. The term tokenization is used in a variety of ways, but generally refers to the process of turning financial assets such as bank deposits, stocks, bonds, funds and even real estate into crypto assets. Under new SEC chair Paul Atkins, the SEC has been attempting to revamp cryptocurrency regulations and reduce rules Wall street has criticized as being overly burdensome. If the latest policies are adopted, they would represent a major win for the digital asset industry, which has long pushed for tailored rules that would enable crypto to become more enmeshed with traditional finance. On Monday, Nasda Nasdaq pointed out that trading of tokenized stocks in Europe is taking place in a way that is raising concerns, as some trading platforms are offering investors access to tokenized US equities but they are not providing investors with actual shares in companies. As part of its new proposal, Nasdaq argued it would raise the bar for tokenized securities to have the same material rights and privileges as do traditional securities of an equivalent class. If those conditions are met, then Nasdaq will trade tokenized securities together with traditional securities on the same order book and according to the same execution priority rules. It exchange will not treat tokenized instruments to be equivalent to their traditional counterparts if they do not convey such rights in whole or in material part, but instead the exchange will treat these instruments as distinct, Nasdaq added. If Nasdaq's proposal is approved, and once the central clearing agency's infrastructure is live, investors could buy a share on Nasdaq and have it settle in token form without changing how orders are routed, price surveilled or reported. End quote.
- [11:08]
B
This episode is brought to you by State Farm Checking off the boxes on your to do list is a great feeling. And when it comes to checking off coverage, a State Farm agent can help you choose an option that's right for you. Whether you prefer talking in person on the phone or using the award winning app, it's nice knowing you have help finding coverage that best fits your needs. Like a good neighborhood, State Farm is there.
- [11:36]
C
When did making plans get this complicated? It's time to streamline with WhatsApp, the secure messaging app that brings the whole group together. Use polls to settle dinner plans, send event invites and pin messages so no one forgets mom's 60th and never miss a meme or milestone. All protected with mirror end to end encryption. It's time for WhatsApp message privately with everyone.
- [12:00]
B
Learn more at WhatsApp.com Eczema isn't always obvious, but it's real. And so is the relief from EBGLIS. After an initial dosing phase, about 4 in 10 people taking EBGLIS achieved itch relief and clear or almost clear skin at 16 weeks. And most of those people maintain skin that's still more clear at one year with monthly dosing.
- [12:21]
D
EBGLIS Librekizumab LBKZ, a 250 milligram per 2 milliliter injection, is a prescription medicine used to treat adults and children 12 years of age and older who weigh at least 88 pounds or 40 kilograms with moderate to severe eczema, also called atopic dermatitis that is not well controlled. With prescription therapies used on the skin or topicals, or who cannot use topical therapies, ebglis can be used with or without topical corticosteroids. Don't use if you're allergic to ebglis. Allergic reactions can occur that can be severe. Eye problems can occur. Tell your doctor if you have new or worsening eye problems. You should not receive a live vaccine when treated with Epglis. Before starting Epglis, tell your doctor if you have a parasitic infection searching for real relief.
- [12:55]
B
Ask your doctor about epglis and visit epgliss.lilly.com or call 1-800-lilly-545-5979.
- [13:02]
E
Abercrombie is an official fashion partner of the NFL and I'm Ceedee Lamb, wide receiver for the Dallas Cowboys. You know I'm here for Abercrombie's Cowboys gear. That's not a question, but I need a whole wardrobe to go with it. No shade to the guys, but I'm used to having the best tunnel fits. This season, Abercrombie has me covered. Shop NFL by Abercrombie in the app, online and in store.
- [13:31]
A
Several interesting raises here. So interesting Raise 1 Remember Mistral, the French based dark horse in the big model AI race? Well, Mistral AI is apparently finalizing a 2 billion euro investment that will value the startup at $12 billion, its first fundraise since June of 2024, when it was valued at $5.8 billion. Quoting Bloomberg, the rival to OpenAI is developing open source language models as well as a chatbot tailored to European users called laychat and other AI services. It was founded in 2023 by Arthur Mensch, a former DeepMind researcher, and ex meta researchers Timothy Lacroix and Guillaume Lample. Mistral has sought to carve out a space in a field increasingly dominated by US And Chinese large language models, and has previously raised over 1 billion euro from investors across the globe, including Andreessen Horowitz, General Catalyst, Lightspeed Venture Partners and BNP Paribas. This is the company's first fundraise since June 2024, when it was valued at 5.8 billion euros. Earlier this year, Mensch said the company was considering tapping outside capital to establish a data center in France. In addition to early funding discussions, the company also recently held talks to raise hundreds of millions of euros in debt for Mistral Compute, its new AI cloud service in France, developed in partnership with Nvidia from French lenders, Bloomberg reported. Well, in light of that, listen to this. ASML is reportedly committing 1.3 billion euro for Mistral's 1.7 billion Series C at a 10 billion euro pre money valuation and so is set to become Mistral's top shareholder and also get a board seat. Quoting Reuters the round will make Mistral the most valuable AI company in Europe with a 10 billion euro or $11.7 billion pre money valuation in its latest Series C funding round, the people said. Often presented as France's and Europe's AI champion, Mistral competes with US giants such as OpenAI and Alphabet's Google. A stake in Mistral would tie together two European leaders, and the cash from ASML could help Mistral make Europe less reliant on US And Chinese AI models, the sources said. ASML makes lithography equipment crucial for cutting edge chip production and uses AI to help make its tools more efficient. The company could benefit from implementing Mistral's data analytics and AI capabilities to improve the performance of its tools and develop additional products. End quote Interesting raise 2 not specifically AI, but listen closely because in a second it is going to be tied into AI. Databricks is reportedly set to close $1 billion in funding at a $100 billion valuation and says it is on track to generate $4 billion in revenue in its fiscal year ending in January 2026, up 50% year on year. Quoting the Journal Databricks, which stores, catalogs and processes large data sets, has reaped the rewards of the artificial intelligence boom. Sales will generate $1 billion of revenue for fiscal 2025, the company said. Data scientists at its client companies use its software to analyze large volumes of information they collect. New customers this year include Honda, Pete's Coffee and Princeton University, chief executive officer and co founder Ali Goetzi said in an interview. Roughly 650 customers currently each pay Databricks $1 million a year for its products and services, he said. Goetze said the company had positive free cash flow in the past 12 months, though he declined to share Oper. We are going to continue to keep this profile where we don't burn any cash on an annual basis, he said. Databricks secured $10 billion from investors last December. Also, the company announced in January $5.3 billion in debt financing from a list of big banks. The latest round of financing will add to the company's war chest for employee recruitment and retention to help it in the escalating AI talent wars. The vast majority of our costs goes to wages and it's definitely expensive to spend on AI talent, goetze said. End quote and finally, interesting raise number three, quoting Bloomberg 11 Labs, an artificial intelligence audio startup capable of creating a wide range of realistic sounding voices, is letting employees sell shares at a valuation of $6.6 billion, double its previous value. The tender offer, which will allow staff to sell as much as $100 million in stock, is an increasingly popular tactic for AI startups engaged in a fierce competition for talent. The offer will let who have worked at the company for at least a year to cash in on valuable shares and allow more investors to boost their stake in the company. Since its founding in 2022, ElevenLabs has attracted both praise and controversy for its hyper realistic voices its tools can create. The company lets users generate audio in multiple languages, create virtual customer support agents, and even make podcasts. End quote first up, quick note that tomorrow is the iPhone event, so tomorrow's show will be a few hours late as usual to account for me watching and telling you all about that event. But more on Eleven Labs and my weekend experiment Longtime listeners will remember ElevenLabs as the AI company I first started using to clone my voice. What was it, three or four years ago now?
- [18:48]
E
Now?
- [18:48]
A
Well, coincidentally over the weekend I read this article about AI generated so called boring history videos. The idea is you get AI to write a script about say the Middle Ages, then you use a soft AI voice to narrate it, put it on YouTube and people use it to fall asleep. In the same way I use audiobooks, just interesting enough to turn your brain off, but not so interesting that you stay awake. These are apparently flooding YouTube right now and the concern is that they are overshadowing authentic historian content, which requires extensive research and production time. Of course. Well, of course I heard about this and I was like I gotta try my hand at it. So Saturday afternoon I prompted ChatGPT to write extensive 10,000 word histories, one of Google and one of Amazon. Then I threw the script into elevenlabs, since I still had an account there, chose a really soft sounding British voice to use, and Voila. I had two roughly 90 minute boring tech history narratives that you could could fall asleep to. So I threw them up on YouTube. The last two links at the very bottom of today's show notes are to those two videos. Check them out. Cause I mean, we're basically there people. We're basically at the point where you can replace Brian the tech historian and tech podcaster with AI. The AI voice is basically perfect. Here are the caveats with this experiment. I did take the time to get it right. I didn't just prompt the AI once and just throw it up without checking, which I'm assuming a lot of the time other video producers are doing. I prompted AI over And over again dozens of times to get things right. You know, okay, give me 800 more words on the founding of YouTube and be more detailed. And I made sure that everything was accurate. I read the whole script, I checked its sources. That's how I knew it was referencing Internet history podcast interviews as sources. Interviews I did a decade ago. So some random thoughts on this. Yes, yes, AI can now do what I do, but is there value in still having me do it? Actually do the AI, me being the curator, me being the conductor of the orchestra, if you will. I know this history, this tech history, so I know what it got right or wrong and. Or I knew when to nudge it to be more detailed or more accurate or more whatever. So sure, the whole thing is AI, but is there value in knowing that it was AI generated by an expert like me in this area? Like, you shouldn't trust me to make a two hour video about particle physics or something I know nothing about. But is there value in getting AI slop if it's been brewed up by someone who knows their stuff? Like, listen to a few minutes of the videos if you seriously wanted to get a fairly comprehensive 90 minute lecture that summarizes the full history of Amazon or Google as companies. I think these videos stand up pretty darn well. More thoughts on this later, but I'd be curious to know what y' all think. This is nowhere near as ambitious as that OpenAI Critters movie that we talked about in the show, but I do think it's quality. Is there a role for AI Slop? If artisans you trust are the ones behind it, you trust them to curate the slop accurately and with taste. I want to do one more experiment along these lines. I'm working on it right now, generating a video on something that I don't know that much about, to see how much work is involved in double checking everything and getting it as right as I can, even if I'm not an expert. More on that when I finish it. But click the links in the show notes, scroll on the videos and listen for a minute or two. Yes, it's AI Slop, but what I'm poking at here is, is it good AI slop? Is such a thing possible? Talk to you tomorrow for the iPhone day.
- [22:35]
E
Packages by Expedia, you were made to occasionally take the hard route to the top of the Eiffel Tower. We were made to easily bundle your trip Expedia made to travel flight inclusive packages are atoll protected.