What's $6B between friends?

Welcome back to Cautious Optimism. Today is May 28th, 2024. A big welcome to new readers from This Week in Startups and This Week in Tech. Let’s have some fun!

Trending Up: Technologist support for (some) AI regulation … neobanks hunting new revenue … wireless consolidation … meme stocks … EV charging stations in the United States … YouTube’s war against ad blockers … iPhone sales in China … vasectomy reversals …

Trending Down: Search traffic for publishers … fintech UX … power reliability in Texas … the ability for the government to quickly build EV charging stations … box office numbers … fusion funding

Earnings Watch: This week we’ll hear from Cava (a 2023 IPO), Box, Salesforce, HP, Pure Storage, Nutanix, Okta, UiPath, Chewy, nCino, C3.ai, Dell, Autodesk, Veeva, Zscaler, SentinelOne, Hashicorp, Asana, and Ibotta (a 2024 IPO).

• It’s a big reporting week for SaaS companies, which are enduring flat+low revenue multiples. Altimeter Capital’s Jamin Ball writes that it’s “still tough out there in software.”

I Dig This Startup: Cloover is a Swedish startup that offers embedded financing to companies selling consumer solar installs. TechCrunch’s Tim De Chant writes that “Cloover recently raised $108.5 million in debt alongside a $5.5 million seed round,” which should give the company plenty of operating capital and funds to, well, fund its offering.

It’s a great example of how a startup can apply a software solution to an IRL market that may help make the planet a little bit healthier. Cloover claims that by offering financing along with an estimate, it can boost “sales and conversion rates by up to 30%.” Thus, more solar panels on more homes. That’s good.

From the grid-level perspective, there’s an interesting economic issue forming when solar installs become so common that energy prices go negative. A rarity for now, but something that we’ll hear more about in the next decade.

Chew On This

This morning, the biggest tech stories center around AI, which I am sure you are shocked by. Two items are most important: xAI’s new, massive funding round and OpenAI's announcement that it is working on its next model.

On the xAI front, the brass tacks are simple enough. Musks’s xAI has raised $6 billion from Valor Equity Partners, a16z, Fidelity, Sequoia and others. Recall that my friend and former colleague Connie Loizos (TechCrunch’s EiC) broke the news that xAI was hunting up double the $3 billion that it initially hoped to raise, and at a higher price to boot. Loizos reported back in April that xAI was hoping to close six billy at an $18 billion price tag. Talk about a scoop.

xAI will compete in a market stuffed with capital, optimism, and big tech connections. Microsoft is besties with OpenAI (while also working on its own AI models); Google and Amazon back Anthropic; Apple is in talks with OpenAI; Meta is taking an open-source-ish approach with its Llama models; Google is working on its Gemini family of LLMs; Oracle, Salesforce, and others back Cohere. Microsoft has put capital into Mistral; you get the idea.

To compete and secure material market share, xAI will need to move quickly and ship products that are at least as good as the best. No small task.

Cautious Optimism: More competition is more good. Though, I have to wonder if we are spending oceans of capital to create models that will be quickly obsolesced by the next generation. Put another way, what fraction of the $6 billion that xAI raised will get spent on models that depreciate quickly? Note that this quibble applies to all frontier-model companies.

What about France? With Mistral and H inside its borders, French startups are keeping the country on the global AI stage today. The two companies have each raised nine-figures worth of funding to date. And yet, if you put their total capital raised together — about $750 million, give or take — you would have to multiply it by eight to get to the number that xAI just raised in one go. France is certainly no AI slouch; but US-based companies can access so much capital, so quickly, you wonder if the EU will manage to build a true AI champion unless it can bolster its fundraising game.

Finally, OpenAI has built a new “Safety and Security Committee,” which feels a bit like a sop to critics after its prior safety team went the way of a sugar cube in hot coffee. More importantly for AI bulls, the company also shared that it “recently beg[a]n training its next frontier model.” GPT5, here we come.

How to ‘save’ Buzzfeed

I’ve kept tabs on work over the past few years to create an alternative, or perhaps parallel internet for folks who think that large tech companies are out to get them. Rumble is one such company, offering video hosting and cloud services to individuals who might not mesh neatly with more mainstream platforms.

Rumble’s first quarter showed 1% revenue growth, and if I am reading its SEC filings correctly, was sharply gross-margin negative in the same period. Off $17.7 million in total revenue, Rumble posted a net loss of $43.3 million. Those are rough numbers. We’ll have to see if its nascent livestreaming and cloud products can improve its growth curve and start to curb losses.

Trump Media and Technology Group is also doing rather poorly, as we recently discussed.

The limited commercial success thus far of companies that want to take a more right-wing bent to The Internet Writ Large is why I was a little perplexed by Vivek Ramaswamy’s recommendations to Buzzfeed, the media company that the former GOP presidential candidate has taken an activist stake in.

As you might expect, Ramaswamy wants Buzzfeed to slash staffing levels. Given its current financial situation, this is not the worst idea, though the scale of cuts suggested appear more like a decapitation than a haircut.

The big idea that Ramaswamy brings to the Buzzfeed table is not a pivot to video — the media company already does lots of video — but instead a pivot to creator-focused video creation with “uncapped upside” for partners coupled with “lower base salaries” to limit Buzzfeed’s required outlays and give it a shot at top-line growth.

You can see why the idea is attractive. Ramaswamy wants to gut existing staff and their output, lever the Buzzfeed brand and existing video subscriber base into a platform for hosting creators’ own work, and eat off its cut of their revenues.

Cheap? Sure. A winning model? I am not so sure. Why? Rumble works with creators of the sort that Ramaswamy wants Buzzfeed to go after — in his words: “From Candace Owens to Destiny, Tucker Carlson to Bill Maher, Aaron Rodgers to Charles Barkley…” — and it has thus far managed to generate lots of losses.

The approach of external investors looking at media brands, writing off their existing teams entirely, and hoping to strip-mine their brand equity continues!

More tomorrow! — Alex