Good morning. Today marks the second inauguration of President-elect Donald Trump. Sam Altman, Mark Zuckerberg, Jeff Bezos, Tim Cook, Sundar Pichai, TikTok CEO Shou Zi Chew and, of course, Elon Musk, will all be present at the event.

In the lead-up to this change of administration, the FTC published a slew of reports related to Big Tech, reports that, under the new administration, will likely lead nowhere.

— Ian Krietzberg, Editor-in-Chief, The Deep View

In today’s newsletter:

• ⚕️ AI for Good: Sustainable cosmetics

• 📱 FTC files complaint against Snapchat’s My AI

• 🚨 Perplexity wants a TikTok merger

• 🏛️ FTC reports on the power of cloud companies

AI for Good: Sustainable cosmetics

L'Oréal said last week that it is partnering with IBM to introduce AI-generated sustainability insights into its operations. This will cover energy reductions, waste reductions and more sustainable formulations for its products. 

The details: The pair plan to develop a custom foundation model — which they believe to be the first of its kind — specifically to address more sustainable cosmetics. 

• The training data for the model isn’t clear — IBM said that it would use a “large number of formulations and component data points” to assist L'Oréal with formulating new products, reformulating old products and scaling up production. 

• L'Oréal said that this partnership will better enable it to meet its sustainability goals, which include a promise that, by 2030, all bio-based ingredients will be both traceable and sustainable. 

"This collaboration is a truly impactful application of generative AI, leveraging the power of technology and expertise for the good of the planet,” Alessandro Curioni, the director of IBM Research, Zurich, said in a statement. "Using IBM's latest AI technology, L'Oréal will be able to derive meaningful insights from their rich formula and product data to create a tailored AI model to help achieve their operational goals and continue creating high quality and sustainable products."

The size, energy profile and carbon footprint of the model remain unclear. 

BILL for your business, Lululemon for you

Want to see how the BILL Divvy Reserve Card can help you get uncapped cash back for your business? Plus get a $200 lululemon gift card for you?¹ 

Get BOTH with BILL Spend & Expense, our powerful combination of the BILL Divvy Card + free expense management software.

Take a 30-minute demo to learn how a secured line with BILL can get you 1.5% cash back with no minimum on eligible business purchases you make every day.²

FTC files complaint against Snapchat’s My AI

The Federal Trade Commission (FTC) has referred a complaint against Snapchat’s My AI feature to the Department of Justice. Such referrals are not normally made public until after charges have been filed, a procedure the FTC chose to forego, saying that its unusual statement is in the “public interest.”

The details: The statement is light on details — the actual contents of the complaint itself were not made public — but the FTC said that it pertains to Snap’s deployment of its My AI chatbot “and the allegedly resulting risks and harms to young users of the application.”

• My AI — powered by models from both Google and OpenAI — was released as a pretty basic chatbot within the Snapchat app in 2023. 

• In investigating Snap for potential violations of the FTC Act, the FTC said that it “uncovered reason to believe Snap is violating or is about to violate the law and that a proceeding is in the public interest.” 

“Since introducing My AI, Snap has harnessed its rigorous safety and privacy processes to create a product that is not only beneficial for our community, but is also transparent and clear about its capabilities and limitations,” a Snap spokesperson told the Hollywood Reporter. “Unfortunately, on the last day of this Administration, a divided FTC decided to vote out a proposed complaint that does not consider any of these efforts, is based on inaccuracies, and lacks concrete evidence. It also fails to identify any tangible harm and is subject to serious First Amendment concerns. 

Considering that a new administration, and a new FTC, will shortly be taking over, it’s not clear if this complaint will actually amount to anything at all. 

• Supercharge your organization's AI adoption journey from complex challenges to strategic success with Tines' must-read guide: "Securing AI in the Enterprise." *

• European Union orders X to hand over algorithm documents (Ars Technica).

• Flights diverted after SpaceX debris falls in path of planes (Bloomberg).

• Lawmakers press Meta, Apple, Google, and others on massive Trump donations (The Verge).

• Google Maps is turning 20 — it’s mapping three more countries and adding AI capabilities (CNBC).

• UK Ministry of Defence enlists sci-fi writers to prepare for dystopian futures (The Guardian).

If you want to get in front of an audience of 200,000+ developers, business leaders and tech enthusiasts, get in touch with us here.

• Apple said it has temporarily disabled its AI-generated summaries for news apps for those using the beta version of Apple Intelligence, a move that shortly follows the BBC’s reporting that the AI summaries were displaying inaccurate information.

• The FTC is taking action against General Motors, accusing the company of collecting, using and selling the precise geolocation data and driving behavior information from millions of vehicles without proper transparency. This data, the FTC said, can be used to set insurance rates.

Perplexity wants a TikTok merger

At the end of a long year of threats and bluster, the campaign against TikTok sort of succeeded.  TikTok briefly shut down the app in the U.S. on Saturday night, shortly following a Supreme Court decision to uphold a law to ban the app unless ByteDance chooses to sell it to an American company. By Sunday, however, the lights were being turned back on, following a promise from President-elect Donald Trump to issue an executive order delaying the ban on Monday. 

And in the midst of this, Perplexity AI — the AI search startup — has submitted a bid to ByteDance for a merger, according to a source familiar with the plans. 

The details: This was first reported by CNBC, who wrote, based on an anonymous source, that the terms of the proposed deal would allow most ByteDance investors to retain their stakes, and would bring more video to Perplexity. 

• Bolstered by the AI boom, Perplexity is now valued at around $9 billion, after starting 2024 with a $500 million valuation. 

• CNBC reported that, since ByteDance seems uninterested in selling the app, the firm might be more interested in Perplexity’s merger offer. It could involve a cash price “well north of $50 billion.” 

ByteDance, however, has made no public indications that it is willing to or interested in selling the app. 

FTC reports on the power of cloud companies 

The interesting thing about the business of AI is the (predictable) power consolidation that the nascent industry has grown up in. It really couldn’t have gone any other way. 

The internet companies of the early 2000s morphed into the social media platforms and the cloud service providers of the 2010s, two essential roles for the past decade of modern history. That dominance supplied those cloud firms with a few things that would later turn out to be really important: capital, computing infrastructure and data. 

When ChatGPT rocketed the field of AI beyond the fringe and into the mainstream business landscape, AI developers found themselves in need of copious quantities of data, enormous amounts of compute and metric tons of capital. 

• The partnerships and acqui-hires that have since proliferated have thus not been all too surprising. The key ones I’m referring to here are the relationships between OpenAI and Microsoft (which invested $13 billion in the startup), Anthropic and Amazon (which has invested $8 billion in the startup) and Anthropic and Google ($2 billion here bought Google 10%+ of Anthropic). 

• This does not, of course, include the deals these three companies have made with startups including Inflection, Mistral, Character and Adept. 

A year ago, the Federal Trade Commission authorized a study into the partnerships between those three leading cloud companies and OpenAI and Anthropic, those two leading AI startups — “these partnerships may potentially impact AI model development — including which firms may effectively participate in the marketplace — and may determine many aspects of those individuals’ and firms’ experiences.”

Here’s what they found: The key pillar of the report is the breadth of the AI tech stack, which encompasses GPU chips, cloud computing, data centers, data and, of course, applications. Microsoft, Google and Amazon are all either developing, or have already developed, AI chips, which, unlike Nvidia’s chips, are only available to those that work directly with those companies. 

All three firms also operate legions of data centers, enabling both the training and operation of developers’ AI models; all three firms have the necessary resources to improve data collection, curation and cleaning and all three firms have numerous pipelines enabling people to access developers’ models (think Azure, Microsoft Office, Google Cloud and AWS). 

• According to the FTC’s report, each of these partnerships involves “significant equity and certain revenue sharing rights in their AI developer partners,” in addition to consultation and exclusivity rights and, perhaps most importantly, “cloud commitments that require AI developers to spend a large portion of their CSP partner's investment on cloud services from their partners.” 

• The FTC referred to this as “circular spending,” adding that key terms of these partnerships include discounted access to those cloud services the developers are required to use. 

These partnerships, according to documents reviewed by the FTC, further gave cloud companies access to “confidential and potentially sensitive financial and performance information,” including customer counts and weekly revenue trends. They have also involved the sharing of model training data. 

The implications: The FTC wrote that these partnerships have the potential to harm competition in the field by limiting access to key inputs, including engineering talent and computing resources. 

“Generative AI requires extensive resources and funding to provide access to infrastructure, access to talent and access to data to train foundation models,” the FTC wrote. “This is why some firms have used technology partnerships, investments and other deals to accelerate development and deployment of AI models.”

The study was published just a couple of days before a new administration is set to take over; current FTC chair Lina Khan, who has made a name for herself by going after Big Tech, will be leaving the Commission. 

🤔 Your thought process:

Selected Image 2 (Left):

• “aww, I was hoping to be wrong — image 1 looked really cool. Ah well, I guess this continues the trend of ‘if there are legible letters, then it's probably real.’”

Selected Image 1 (Right):

• “I was 51/49 on this one. In the real image, I didn’t like the clean cut of the dirt on the turn — seemed too precise.”

💭 A poll before you go

Thanks for reading today’s edition of The Deep View!

We’ll see you in the next one.

Here’s how aware you are of your daily AI interactions:

37% of you are very aware of your daily AI interactions (half of them use chatbots regularly, and half don’t). 14% are somewhat aware, but don’t think about it too much; 17% were not aware at all.

Very:

• “Candidly, I would expect anyone reading this email would be in the ‘very’ category. Hmmmmm……”

Me too. Me too. But it looks like not quite everyone is there.

_{*BILL’s disclosure:}

_{¹Terms and conditions apply: see offer page for more details.}

_{²Rewards are subject to your Rewards Agreement. For more information on Rewards see here.}

_{The BILL Divvy Reserve Card is issued by Cross River Bank, Member FDIC, and is not a deposit product.}