Plucking handcrafted signal from the roaring stream [Graphic via Gemini]

What a year this week has been. When even my kids are asking me about the Silicon Valley news, because it's all over their TikTok, you know the headlines have reached escape velocity. With deference to Nvidia as an exceptional success story, a central player to the massive AI boom (not to mention partner to my company and pretty much most), seeing a leader go from a universally adored to a memed and mocked punchline can give anybody whiplash. Did the individual change in the last week? Probably not. But what we know has, and there are a lot more questions that will keep this story going for a while.

It was just last Friday when the Financial Times led with the headline of the year: "US believes Nvidia’s Huang had romantic relationship with alleged Chinese spy." This is the kind of story that journalists dream about. I can imagine the incredulity when the whispers turned to rumors, and the nervousness of vetting all the sources to be sure you had the right story, getting the blessing to publish from the editor (and legal, no doubt) and then hitting publish. That turned out to be the hole in the dam, which led to a week of leak after leak that flooded the zone.

Good luck trying to get attention for your software launches this week when all people wanted to talk about was the CEO's downfall, what preceded it and what may come afterward. But, the innovation must continue. MG Siegler captured the dizzying feeling well in AI’s Ringmaster Loses the Circus over at Spyglass. Longtime readers of this blog will remember MG and I grew our tech audiences in parallel, and he eventually joined me and worked at Google Ventures for a good part of a decade. While his outlet has changed a few times, his industry observations continue unabated. Subscribe to Spyglass right now if you want his independent commentary.

Drama at the top of the chip giant's org chart was only the most noisy piece of a very crowded calendar this week. Over at Chrome Unboxed, the team had early feedback to Google and Samsung’s smart glasses. I'm happy to say my corporate avatar still features a photo of my wearing the original Google Glass from I/O 2013. They, like many launches in Valley history, were ahead of their time. Now it seems all the top vendors are getting back into this space, including Meta and Apple.

Also catching my eye this week was Ben Werdmuller's piece on the shifting media landscape, where he wrote that News needs entrepreneurs, asking Where will they come from? on Werd.io, challenging us to support the builders who refuse to rely on algorithmic distribution. In that article, he says "Journalism is operating in an existential threat environment," adding "We are in the fastest period of technical change in my lifetime." There's no doubt this is the case. When I got my degrees in Mass Communications and Political Science at Berkeley, the Internet was already changing how we discovered and shared content, but it was all so new, and the biggest newspapers still ruled the land. That's a forgotten era now.

Sifting through this digital deluge is a daily labor of love, and I've recently added deeper sharing integrations to the new Webnerd News, which follows on the launch of the Webnerd Chrome Extension to accelerate the amplification of links across the web. It should be exceptionally easy to consume and discover the best of the web. Back in 2013, I wrote in Shared Items Might Be Gone, but the LG Stream is Back that acting as a human filter to highlight the best corners of the web has always been my favorite way to engage with technology. It's still true.

You can follow along with everything I am reading and sharing by checking out my public shared stream directly, or by plugging the dedicated RSS feed into your reader of choice. Let us keep championing the independent voices that make the open web worth exploring.

Disclosures per always: I work at Google. Nvidia, Meta, Apple, Samsung and many companies mentioned here are partners, competitors or customers, and I have plenty of friends working at each place. Don't interpret my interest as being filtered with that lens. Also, I built Webnerd.ai (and have no plans to monetize it).

Webnerd News (in Labs) brings visual clustering to top stories and streams.

An algorithm-free stream of updates, sorted chronologically, ensures you never miss a single post. Bringing that classic RSS reader feeling back, including recommended bundles and sharing, has reinvigorated the discovery loop for me after a dark decade dominated by social feeds fueled by ragebait, attempts at viral videos and headlines that left you questioning what stories were all about. But there's also value from automated or hand-selected curation you see from sites like Techmeme and major news hubs like CNN or Google News.

Typically, these sites are fed by their own first party sources or editor curated content. As a viewer, they may slightly adjust based on their incremental levels of personalization, but for the most part, you get the same view everyone else does -- and your headlines or top stories are the ones they see.

What if you could combine the at a glance view of what's most important, including story clustering when many sources cover the same topic, but built on a foundation of the sources you choose to read? That's the idea with a new view in Webnerd.ai -- a fun experiment that combines the feeds you personally subscribe to, augmented by the starter packs I engineered into the site from the first day, ensuring you always have something to read, and constantly have new sources to discover and subscribe to.

Building on the web platform, and offering a direct connection from the publisher to the reader is an important foundation of what makes RSS special. If everything is a feed, you can point that feed and shape it every which way. We remember this from the 2010 era wave that saw Flipboard unveil its social magazine on the iPad alongside Feedly, Plaxo Pulse, Ninua and many others, even as plenty of pundits rushed to declare classic feed readers dead. I pushed back at the time saying RSS or RIP, the Results Always Trump the Methods, arguing that consumption tools aren't a zero-sum game. Sometimes you want the raw river, item by item. Other times you want an editorial front page that groups the noise into a coherent briefing. After watching filtertainment eviscerate our informed monoculture over the past decade, building both modes on top of open feeds feels overdue.

So let's have fun with it.

Clustering Your Subscriptions With 14 Starter Packs

Tapping Full Coverage lines up reporting, deep dives, indie bloggers, video, and podcasts on the same story.

The engine underneath Webnerd News starts with every blog, newsletter, YouTube channel, and podcast you have connected, then blends in active coverage across all 14 Webnerd.ai Starter Packs. Strict source diversity caps keep any single high-volume site from flooding the page, and if you still haven't tried Webnerd.ai yourself, and take a look without logging in, it defaults to an edition generated from my own subscriptions so you can kick the tires right away.

Instead of repeating duplicate headlines, related posts converge into a single story cluster tagged by the role each piece plays, offering:

• Perspective Tags & Full Coverage: Every item carries a clear badge -- Reporting, Deep Dive, Indie Take, Community Pulse, Video, or Podcast. Hit Full Coverage to compare every angle side by side or step through a chronological timeline of how the story unfolded.
• Inline Starter Pack Discovery: When a writer from a Starter Pack you don't follow lands inside a story cluster you're reading, or surfaces in the Discover New Voices rail, you see which pack brought them in alongside a one-click + Follow button that subscribes and files them straight into a matching folder.
• Zero Unread Anxiety: Browsing Webnerd News is strictly read-only against your classic reader queue. Flipping through the magazine never marks articles read behind your back, and stories you already read earlier in the morning still stay on the board for context

This is very much day one of a broader buildout and another fun volley in vibe coding to victory. I plan to keep tuning the clustering weights, tightening how personal folder signals steer the front page, and expanding the Starter Pack catalog beyond tech and media. Take Webnerd News for a spin on your own feeds today and let me know on Bluesky what clicks, or what you'd love to see next.

Product vision is as valuable as coding excellence [Image via Gemini]

As long as I've been in Silicon Valley, engineers have been king. The most sought after, and highest paying, roles have gone to the code orchestrators with the highest talent, who have been able to turn ideas into reality. Expertise could be claimed due to the number of coding languages they knew, and effort could be arbitrarily counted by the number of lines of code, with only its complexity up for debate. The most flexible and most capable engineers have deftly adjusted their expertise to meet the industry's latest demand, from dedicated apps for the top operating systems at the time, to mobile applications, social network tools, ride sharing apps, autonomous driving, crypto, or whatever business fad or innovation has driven the newest industry.

But now we're in an odd spot. Many of the software engineers I know and have worked with for years are telling me they scarcely code at all, with AI-assisted coding (vibe coding if you must call it that) making it possible for them to tell agents what to do and how to build at unprecedented scale. Meanwhile, recent college graduates sporting computer science degrees are finding entry level opportunities have practically disappeared.

Every technology shift has downstream impact, and the seismic shift toward AI-driven tooling means the roles in demand in this new world have evolved. Lenny Rachitsky's breakdown of the tech job market puts hard numbers behind a shift many of us have seen with our own eyes or felt when looking at our in boxes and LinkedIn messages. Demand for traditional software engineers is plateauing, open mobile and front-end roles have been nearly cut in half, and now product managers are suddenly the ones seeing the biggest hiring surge. 

When I wrote about vibe coding acting as automatic transmission for apps last month, the takeaway wasn't that humans were obsolete, but that the mechanical act of shifting gears was no longer the bottleneck. If a coding agent can build out an app or deliver a mobile view in an afternoon, paying a premium for narrow implementation stops making sense. As Gergely Orosz pointed out in response to Lenny (see Techmeme), specialists who only know one slice of the stack are getting squeezed out in favor of generalists who can steer AI across the whole system.

There is no such thing as a sure thing in a turbulent time. Back in 2018, I looked at Silicon Valley's divergent career arcs and how quickly comfortable middle-tier roles can evaporate when the industry's priorities tilt. We are watching that exact thing play out today in real time.

Telling a machine what to build requires knowing what is actually worth building. If code generation gets cheap, product judgment becomes the scarce resource, which is why product managers and customer-facing builders are having a moment, living in the messy human problem space where no model can hallucinate real user empathy. As I argued in my October piece on human evolution in an AI revolution, the winners won't be the folks guarding the roles and tools of yesterday, but those who are flexible enough to see how the world is changing and be part of that movement themselves.

All software has bugs. It has always been that way, and much of the recent discussion around AI's impact on software engineering has been about unveiling security holes that plague popular software and may have sat for years -- hardly the setup for promising complete code verification.

I saw the headlines around DeepMind's AlphaProof Code announcement this morning, saying they confirmed Signal's security protocol to be solid, and initially glanced past it, as someone who typically doesn't use Signal. But the post has ignited a foundational discussion about how software will be built and verified going forward in the usual places, including Hacker News, which has zeroed in on a major turning point: moving from classic test-driven development to automated, machine-driven formal proof generation.