Funston Brief (Issue No. 20)
This month, our featured commentary discusses the antitrust case against Google. We also dive into the latest developments in AI models and LLMs.
The Justice Department is mounting its biggest antitrust challenge against Google, focusing its allegations on TAC (Traffic Acquisition Cost) payments. Google’s search engine controls almost 90% of online queries. The US government contends the tech giant has paid billions of dollars to maintain a monopoly over the search market. However, it’s unclear whether when Google pays Apple $20B a year to be the default search engine on iPhones, they are abusing their search share or whether Apple is taking advantage of its handset share. In a quiet move, John Giannandrea, a former top Google executive now heading Apple’s AI business, pointed out a recent change in the latest iPhone software update that allows users to select a search engine other than Google. The Google antitrust trial isn’t just one of the biggest antitrust cases of the decade - it’s also one of the most secretive, although the judge said documents used during the trial can be published online at the end of each day. Other big tech companies like Apple and Microsoft are joining Google to limit public access to the trial. Regardless of the outcome, the Google trial will be another test of some US regulators’ project to persuade courts to take a more expansive view of US competition law. In another recent move, the FTC and 17 states sued Amazon earlier this week, alleging that the company holds a monopoly over online retail, artificially inflating prices and restricting sellers’ choices.
In other legal news, the SEC has recently had a run of court losses after spending most of its past year challenging crypto projects that behave like unlicensed securities. Although most individual crypto tokens are often unlicensed securities, it’s been hard for the SEC to apply the 1946 ‘Howey Test.’ The SEC is trying to devise new regulations to apply to the crypto world.
ACTIONABLE INSIGHTS.
Open-source AI models: Meta continued its open-source AI push, releasing a version of Llama for coding. As experimental budgets increasingly pour into ChatGPT, OpenAI launched an enterprise version that boasts security, privacy, and SSO. At the same time, Google Cloud will provide models from both Meta and Anthropic on Google Cloud.
Key Insight: Coding is one of the most apparent early use cases for LLMs, and Microsoft wants to monetize this. In other news, Microsoft also announced a new computer vision model and, interestingly, a benchmark for assessing the fairness of image recognition models.
LLM as a feature: Bard, Google’s search companion and its competitor to ChatGPT got an upgrade that allows the artificial intelligence chatbot to connect to a user’s Gmail, Google Docs, and Google Drive accounts and use it across any or all of them. Microsoft is replacing Cortana with Copilot in Windows, and Amazon is replacing Alexa with a generative version.
Key Insight: The buzz around consumer generative AI has died down since its early 2023 peak, but Google and Microsoft’s battle for AI supremacy may be heating up again. New technology is usually embedded into the old products in every platform shift. The next and most exciting step is to create new products/apps native to the latest technology. In a recent blog post, Sequoia Capital argues that we are now entering that phase - Generative AI’s second act. Link Link Link
EAR TO THE GROUND.
Podcast Episode: Bloomberg Originals - Exponentially With Azeem Azhar
Guests: Niall Ferguson
Takeaway:
Is humanity living through a historic transformation - a paradigm shift in how societies organize themselves, their economics, and politics? Or are we traveling through another manifestation of the trials and triumphs of those who came before? On this episode of the Bloomberg Originals series Exponentially, Azeem Azhar talks with Hoover Institution fellow and historian Niall Ferguson about whether civilization is entering a new era.
“Unprecedented things are actually quite common in life, and most of the things that are happening in this particular year aren't drastically different from the things that struck people as worrying fifty years ago or a hundred years ago.”
- Niall Ferguson, Historian and Hoover Institution Fellow
$4.87B
Arm Holdings raised $4.87 billion in the largest listing of the year, delivering a signal that the market was ready to reopen to fresh listings. Arm had marketed 95.5 million shares for $47 to $51 each. Despite being oversubscribed more than ten times, SoftBank left millions on the table - pricing the IPO just $1 more a share would have raised about $100 million more. After climbing 25% in its trading debut, the chip designer fell below its IPO price just one week later.
PROFILE.
Toto Wolff built Mercedes’ F1 team into an auto racing dynasty. Over the past decade, Wolff has amassed eight Constructors’ Championships, seven Drivers’ titles, and 115 Grand Prix victories thanks to his flawless cars and Mercedes’ marquee driver, Lewis Hamilton. Dive into his winning formula here!
“Choosing between financial success or sporting success, every day of the week, every day of the year, I'll go for the sporting success.” - Toto Wolff, CEO of Mercedes F1 Team
ON MY RADAR.
OpenAI and Jony Ive, the famed Apple designer, are in talks to raise $1 billion from SoftBank for an AI device venture, the "iPhone of artificial intelligence." Link