Tech Brew Ride Home - Thu. 08/01 – Now Meta’s AI Spending Is Fine?

Episode Date: August 1, 2024

So I guess Zuck has convinced Wall Street that all the capex spending on AI is fine. How good a business is Tether? A gooooood business. The first arrival of Apple Intelligence is sort of a ho-hum for now. And how sports is completely transforming the streaming wars battlefield. Links: Meta’s Lesson in Allaying Wall Street’s A.I. Fears (Dealbook/NYT) Meta’s AI Costs Surge as Digital Advertising Revenue Grows (WSJ) Reddit CEO says Microsoft needs to pay to search the site (The Verge) Tether Nets $5B Profits This Year, Says Its US Debt Holdings Surpassed Germany's (CoinDesk) A first look at Apple Intelligence and its (slightly) smarter Siri (The Verge) Disney, Fox and Warner Bros. Discovery’s Sports Streamer Venu to Launch at $42.99 per Month (Variety) How the NBA’s Slam-Dunk Deals Will Reshape TV (The Hollywood Reporter) Learn more about your ad choices. Visit megaphone.fm/adchoices

On April 4th, 2023, around 2 in the morning, a man was found stabbed multiple times on a sidewalk in downtown San Francisco. Hey, who did this to you? What happened next turned the story into a political firestorm. Reports have identified the victim as Bob Lee, the founder of Cash App. From Bloomberg Podcasts, this is Foundering, the Killing of Bob Lee, beginning April 16. Welcome to the Tech meme right home for Thursday, August 1st, 2021. I'm Brian McCullough today. So I guess Zuck has convinced Wall Street that all the CAPEX spending on AI is fine.

How good a business is tether? A good business. The first arrival of Apple intelligence is sort of a ho-hum for now, and how sports is completely transforming the streaming wars battlefield. Here's what you miss today in the world of tech. Meta's stock is up this morning in pre-market trading after reporting Q2 revenue up 22% year-on-year, net income up 73% year-on-year, and the family daily active people number was up 7% to a total of 3.27 billion. But the interesting thing here is that the stock is up,

even though the Reality Labs unit is still losing money, now to the grand total of around $50 billion since 2020. And guess what? The spending on AI research and development will continue. So what appears to have happened here is that Zuck has somehow found a way to explain away all the things that had been concerning investors in recent years, at least in a way that has actually gotten them on board with his spending again. Quoting the New York Times' deal book from this morning. Another tech giant has run the gauntlet of investor worries about the costs of commercializing artificial intelligence, except that this one passed with flying colors. Shares and meta are up more than 5% in pre-market trading, even after the parent of Facebook and Instagram said it had increased spending on AI and other projects by 33% from a year ago. That suggests that while shareholders remain worried about whether such mammoth investments will pay off,

they're willing to be patient if a company's core business holds up. The figure in focus was capital expenditures with meta spending $8.5 billion in the second quarter. That also includes the company's long-running unprofitable investments in the so-called Metaverse and other endeavors. CEO Mark Zuckerberg made no apologies about the big spend, quote, I'd rather build capacity before it is needed rather than too late, he told analysts on Wednesday. Underscoring the point, meta-executive said they expected capital expenditures to reach at least $37 billion this year, up from $35 billion, but no more than $40 billion. That's in line with what Microsoft and Alphabet said about their AI spending, but exceeding expectations in core operations helps,

as opposed to falling short, however, slightly. So, too, does Meta's point emphasizing that AI innovations would bolster its ad business by showing users more relevant content, end quote. Right. Meta makes its money on ads, and it believes AI will make for better ads, quoting the journal. Meta's shares have risen by almost 50% in the past 12 months, a rally fueled by breakthroughs and artificial intelligence technology that have improved its ad-targeting capabilities. Those advancements in AI have helped meta overcome challenges. posed by privacy changes implemented by Apple in 2021 that raised $10 billion of revenue for the social media company in 2022, end quote. On the earnings call, Zuck said he can see a day in the near future

where advertisers can just explain a broad business goal and meta's AI will take care of the rest, creating the ads, flighting them, tweaking them, optimizing them. Sort of feels like a holy grail if again you are largely an advertising business. So great, investors can look past all the AI spending because they think there will be tangible benefits soonish. But there's also this. On the earnings call, Zuckerberg said Lama 4 will likely need around 10 times more compute than Lama 3.1 did to train, and future models will be bigger even beyond that. Now, this lines up with recent comments by Anthropic CEO Dario Omadi, who said each new foundational AI model tends to be 10x the previous one. We've already seen, we believe, the first 100

million-dollar models. We assume the first billion-dollar models are being trained right now, likely by meta, among others. So number one, roll up the Brinks truck if you're Nvidia, but number two, you can see why the major model makers are throwing caution to the wind right now. If in two years' time, you need $10 billion just to train one state-of-the-art model, better to have what you can now out in the world while people are willing to give you the money to build it, because the numbers there suggests that one day there could only be the likes of meta and Google. and Amazon at all that could afford to even train the state of the art. Kind of makes you see why Zuck wants to go open source.

But also, beyond those numbers, are we getting into the territory of my suggestion that maybe nation states will have to foot the bill to train models once the costs reach the hundreds of billions of dollars or more? Following up on something we've been discussing, Reddit CEO Steve Huffman, says Microsoft Anthropic and Perplexity need to pay up to scrape Reddit's data because blocking them recently has been, quote, a real pain in the ass. Quoting the verge. After striking deals with Google and OpenAI, Reddit CEO Steve Huffman is calling on Microsoft and others to pay if they want to continue scraping the site's data. Without these agreements, we don't have any say or knowledge

of how our data is displayed and what it's used for, which has put us in a position now of blocking folks who haven't been willing to come to terms with how we'd like our data to be used and not used, Huffman said in an interview this week. He specifically named Microsoft Anthropic and perplexity for refusing to negotiate, saying it has been, quote, a real pain in the ass to block these companies. Reddit has been escalating its fight against crawlers in recent months. At the beginning of July, its robots.com file was updated to block web crawlers. It doesn't have agreements with. Then people began noticing that Reddit results were only visible in Google results, where Reddit is paid for its data to be shown, and not other search engines like Bing.

Huffman said that Microsoft has been using Reddit's data to train its AI and summarizing its content in Bing results, quote, without telling us, and that Reddit's data has also been sold through the Bing API to other search engines. In the interview, he referenced Microsoft AI CEO Mustafa Sullyman's recent comment at a conference that public data on the internet is, quote, freeware. We've had Microsoft Anthropic and perplexity act as though all of the content on the internet is free for them to use, Huffman said. That's their real position. After the story was published, Anthropic spokesperson Jennifer Martinez sent the following statement. Reddit has been on our block list for web crawling since mid-May, and we haven't added any URLs from Reddit to our crawler since then.

We respect robots.com. The industry accepted signal for blocking web crawling, end quote. Microsoft declined to comment for the story. Perplexity didn't respond to a request for comment either, end quote. I believe I've suggested in the past that running a stable coin is a good business, but even I didn't know how good. Tether, which is a type of stable coin, though maybe not exactly a stable coin, don't at me, has reported a record $5.2 billion in net profit for just the first half of 2024, a net operating profit of $1.3 billion in Q2, and $97.6 billion in U.S. debt holdings, which would be more than countries like Germany. So crypto as a shadow global finance force, it's already happened, people,

quoting Coin desk. According to its latest quarterly attestation signed by accounting firm BDO in Italy, the company's stablecoin issuer arm Tether International Limited and Tether Limited disclosed $118.4 billion of assets in reserve against $113.1 billion in liabilities as of June 30th. This translated to $5.3 billion of excess reserves backing Tether Stablecoins. Its U.S. debt holdings amounted to $97.6 billion, which would rank as the 18th among countries surpassing Germany. The United Arab Emirates and Australia, the company said. Tether investments, the entity established as a separate division from the stablecoin business to manage the companies growing four-way into Bitcoin mining, artificial intelligence and other investments,

has a $6.2 billion net equity value per the attestation. Tether's USDT is a key piece of infrastructure in the crypto market for trading and is increasingly in demand in developing countries as a vehicle to access the U.S. dollars. It's the most popular stable coin with $114 billion market value, up from 91, billion dollars this year per coin gecko, end quote. Apple rolled out iOS 18.1 betas with some of the Apple intelligence stuff arriving. What's the consensus on this stuff so far? The verge says there are meaningful Siri language understanding improvements and Siri text input is great, but there are a few other notable features, at least right now.

Quote, the big Siri AI update is still months away. This version comes with meaningful improvements to language understanding, but future updates will add features like awareness of what's on your screen and the ability to take action on your behalf. Meanwhile, the rest of the Apple Intelligence feature set previewed in this update feels like a party waiting for the guest of honor. That said, series improvements in this update are useful. Tapping the bottom of the screen twice will bring up a new way to interact with the assistant through text. It's also much better at parsing natural language, waiting more patiently through hesitations and ums as I stumble through questions. It also understands when I'm asking a follow-up

question. Outside of Siri, it's kind of an Easter egg hunt finding bits of Apple Intelligence sprinkled throughout the OS. They're in the mail app with a summarized button at the top of each email now, and anywhere you can type and highlight text, you'll find a new option called writing tools with AI proofreading, writing suggestions, and summaries. Help Me Write Something is pretty standard fair for generative AI these days, and Apple Intelligence does it as well as anyone else. You can have it make your text more friendly, professional, or concise. You can also create summaries of text or synthesize it into bulleted lists of key points or a table. I'm finding these tools most useful in the Notes app where you can now add voice recordings. In iOS 18, voice

recordings finally come with automatic transcriptions, which is not an Apple Intelligence feature, since it also works on my iPhone 13 Mini. But Apple Intelligence will let you turn a recording transcript into a summary or a checklist. This is helpful if you want to just free associate while recording a memo and list a bunch of things you need to pack for an upcoming trip. Apple Intelligence turns it into a list that actually makes sense. But despite the light show, Siri is about the same as ever. It mostly remains a let me Google that for you machine. The most significant updates are still to come in future updates when Siri will gain awareness of what's on your screen and will be able to take action in apps for you. In theory, you'll be able to have Siri grab

information from messages and turn them into calendar events or retrieve information from email without you having to go digging through your inbox. That's the stuff I'm most excited about, and all of the pieces of Apple intelligence available so far could be the building blocks of a better Siri. Apple's AI is capable of understanding the contents of an email or a photo. Likewise, Siri is better at understanding how humans talk. In order for Apple intelligence to prove itself, Siri needs to connect the dots, end quote. Finally, today, there's a big new streamer coming for us that I want to remind you about. Venu is a joint venture between Warner Brothers Discovery, Disney, and Fox. It will launch in the fall, we believe,

and it will contain all of the sports assets of those companies. We now know it will cost $42.99 per month, at least at the start. Quoting variety. Venu plans to burnish the sports offerings of 14 different TV networks, as well as a library of documentaries and other programs from the archives of ESPN, Fox Sports, and others. Subscribers to Venu will have access to linear sports networks including ESPN, ESPN, ESPN News, ABC, Fox, FS1, FS2, BTN, T&T, TBS, True TV, as well as ESPN Plus. The price has long been expected to total more than a consumer would pay for a standalone regional sports network, which costs $20 to $30 per month, and less than a larger digital programming package such as Hulu Plus Live TV or YouTube TV, which costs around $75 to $80 per month, end quote.

Yeah, I would expect that initial price to go up quite quickly, so if you're interested, jump now. But I wanted to mention that story to also mention this. At this stage in the streaming wars, as the streamers turn their heads to sports, that is beginning to crowd everything else out. The Hollywood reporter says, Sports Rights Deals are moving money out of Hollywood

and into the coffers of the major sporting leagues. Sports now make up 40% of Disney's content budget, and as an example, NBC's new NBA deal will replace more than 150 hours of original show programming. Quote, The NBA's Blockbuster $76 billion mega deals with Disney, NBC Universal, and Amazon Prime are a testament to its product, and the unrivaled reach and power of live sports.

And the deals should also send shivers down the spines of almost everyone in Hollywood outside the network executive suites. While the 11-year agreements are seen as a critical strategic maneuver by the media companies looking to build streaming businesses and are a financial boon for the league, they are also, in the words of one veteran media executive who declined to be named, the latest example of a, quote, transfer of wealth from Hollywood to the sports leagues. Already agents and writers are grumbling that the rich deals will mean fewer new TV shows ordered and fewer reruns, which will result in lower residuals. NBC's deal alone will replace more than 150 hours of broadcast TV entertainment with live NBA programming on Sunday and Tuesday nights,

not including the playoffs, NBA All-Star Weekend, or select WNBA games, which will also run on NBC. When you zoom out and think about the total picture of what we're trying to do, which is to bring our excellent TV media assets into the future, I think we view the NBA as an excellent piece in that puzzle. Comcast President Mike Kavanaugh told Wall Street analyst July 23rd, while adding that quote, it will allow us to rebalance programming from other areas. Obviously, we'll fill a few nights on NBC with this content versus other content, end quote. That other content will be largely entertainment programming. NBC's schedule this year programmed new episodes of The Voice and Nightcourt on Tuesdays, among other unscripted and scripted shows. Sunday nights after the NFL season ends have been a home

for reruns of NBC scripted shows or movie nights. The NBA deals arrive as the entertainment business is already in a state of contraction. TV networks and streaming services have been rationalizing their budgets, which are either flat or increasingly down, as everyone except Netflix seeks to make unprofitable streaming businesses into something that can actually make money. Netflix, meanwhile, seems content to keep its content budget more or less flat, but is spending more of it on live events, be it low nine figures for a pair of Christmas NFL games, or $500 million per year on the WWE. Across the world of entertainment, sports appear to be eating up more of the overall budget. Disney said that it expects to spend $25 billion on content this year

down from $27 billion in 2023. 40% of its content budget is dedicated to sports and sports-adjacent programming. There's less money overall and more of that money is being allocated towards sports, says Jonathan Miller, a former NBA executive, who serves as CEO of Integrated Media, which specializes in digital media investments. The sports audience is a more or less guaranteed audience, predictable. You can sell against it. You kind of know where it's going to fall within ranges, end quote. The NBA deals will only make the spread between sports and entertainment budgets more obvious. The new deals will see NBCU pay about $2.5 billion per year to the NBA, not including the cost to produce the live games and pre- and post-game shows, known as shoulder programming,

while Disney will go from paying $1.4 billion per year under the previous deal to about $2.6 billion under the new deal. Suffice to say, few in Hollywood expect either company's overall content spending to increase by the same margin, even if they do rise slightly. That said, at least for NBC, executives hope the NBA can also bring a jolt to its entertainment lineup. In a memo to NBC staff after the deal was announced, Kavanaugh wrote that, quote, the league reaches a broad, diverse, and youthful audience that is culturally relevant, end quote. I forgot to tell you that I sheld out for a Bukes Palma, that handheld e-reader in the form factor of a smartphone, thought I'd give it a go as just that as a Kindle alternative, and it lasted all of 12 hours before the screen broke, or got all artifact-y, at least it's unusable. It was already somewhat annoying to use, because it didn't do simple things like, remember the brightness setting I wanted from one time turning it on to the next, and I didn't even get to install any apps beyond my Kindle app before again, screen busted.

So back it goes, which is a shame, because I was excited. about that as a little experiment. Talk to you tomorrow.