The quiet before MS Build

With Google I/O in the rearview mirror and Microsoft Build looming large next week, this was a relatively quiet week. Now that Microsoft has emerged as the unlikely leader in the AI race (thanks to OpenAI), you can guess what next week will be all about: AI copilots everywhere.

Microsoft decided not to invite the press to the in-person event (which is weird, given that they were insistent that everybody come to Redmond for the Bing announcement). I admit, that makes it much harder to care about the event.

When virtual conferences became a thing during the pandemic, I watched the keynotes for a few and quickly burned out on them. Typically, I have all of the news ahead of time anyway, so being at an event is all about talking to people, be those execs, vendors or attendees. That's the only way to really get a feel for what's happening in a given ecosystem outside of the quickly deflating Twitter bubble (remember when conference hashtags used to trend on Twitter?). The hallway track is always the best part of every conference. Without that, I'll just read the news.

Open Source businesses

My stories this week mostly focused on a number of funding rounds for open source companies -- one of them led by Craig McLuckie, the founder of the Kubernetes project, and Luke Hinds, the founder of the sigstore project, which now underlies most software supply chain services. Given their background, that's an easy round to cover -- and I've always enjoyed chatting with Craig ever since I covered the original Kubernetes announcement back in 2014 (at the time, I thought of it more as a complement to Docker. Little did I know...). Craig teamed up with his fellow Kubernetes founder Joe Beda to launch Heptio and then sold that to VMware, so it's also no surprise that some of the same VCs that backed Heptio now returned to back his new startup, which focuses on supply chain security.

What was interesting this week, too, was that virtually every company I talked to about funding this week (that's Stacklok, Union AI and Visual Layer) followed what's essentially the same playbook: the co-founders launched successful open-source projects and now they are monetizing them, typically with plans to offer hosted services of some kind. That's almost a meme at this point. In the world of developer tools, these open-source projects are now basically MVPs. Often, founders start these projects inside large companies. The companies then often spin those projects out, because, for the most part, developers are suspicious of open-source projects that are mostly backed by a single industry giant. Having built a successful project that's likely attracted some attention from VCs, the founders then often leave that big company and raise some money for their startups.

Since these projects often solve very specific problems, we'll have to see how this model will play out in the coming months, as businesses start to look into reducing their expenses (and, in the process, often consolidate where they spend their money).

Cookies

Also this week, I talked to Victor Wong at Google, where he leads product for Private Advertising Technology within Privacy Sandbox. Starting early next year, Google will deprecate third-party cookies for 1% of Chrome users and move them to Privacy Sandbox, with plans to move everybody away from third-party cookies by the second half of 2024. That's a massive change for the ad tech ecosystem, even if the other browser vendors remain on the sidelines and often favor their own approaches to blocking trackers.

Google's promise is that online publishers will be able to use its various new APIs in Privacy Sandbox to still target ads, but based on the kind of topics users are interested in (which the browser will infer from your browsing habits and which you will have control over).

To call all of this controversial would be an understatement, but Google is also being closely watched by regulators, especially in the U.K., to avoid any self-preferencing. It also argues that Privacy Sandbox reduces the incentives for ad tech companies to invest in browser fingerprinting and other techniques that users have no control over.

The thing is, I can see where Google is coming from (and online ads pay a good chunk of my salary), but I can also see why many would prefer alternatives that weren't developed by the world's largest online advertising company. In the ideal world, nobody would track you online. Privacy Sandbox, to me, feels like a reasonable alternative that, if executed right, protects a user's privacy and if the ad tech companies don't like it, that's probably a good thing.