Silicon Valley Technology Commentary & Archives · Est. 2006 3,053 Posts · 2006–2026

November 19, 2026

November 19, 2026 · 4 MIN READ · BY LOUIS GRAY

Nerdticker Blends Tech Curation With Bluesky Voices

Nerdticker Blends Tech Curation With Bluesky Voices

Nerdticker two-column front page showing Top Stories and Bluesky Reactions with outlet handles, Follow and Share actions, and The Ticker live wire
Nerdticker pairs clustered tech headlines with live reactions and shares from Bluesky.

As I mentioned in yesterday's post about Webnerd News fighting silos with a focus on feeds, follows and shares, I feel we have a responsibility to the open web to encourage discovery, sharing and attribution. We've lived through the evolution that saw news move offline to online, for individual authors to gain prominence from their individual outposts (blogs, newsletters, channels and so on) and then watched as social networks absorbed a lot of this momentum and learned to keep users on their streams instead of on the broader Internet.

The launch of webnerd.ai last month brought back the classic RSS feel to the modern web. Webnerd News took that foundation of your feeds, augmented it with starter packs, and built out a visual representation of what the top stories are, while bringing following and sharing to the forefront, explicitly for those of us who act as human filters, curating the best of the web. The third pillar is deeply integrating this motion with the broader community, focused on the social network that is most open to full links and the open graph -- Bluesky, which also happens not to be tainted with the toxic dreck found on other networks.

Nerdticker blends the best of Webnerd News with the best of Bluesky and surfaces conversations and shares from the network to show who is talking and engaging on the tech (and related) news of the day, and true to form, highlights the same follow and share capabilities we need in a healthy web ecosystem that's been gathering dust. After years of seeing so many of my favorite services I used to discover, share and amplify go away, I've decided to build my own. Vibe coding with Antigravity has made this possible.

Nerdticker should look familiar, dressed in the same scheme as my blog, featuring ranked story clusters, Bluesky handles, inline Follow and Share actions, and live Bluesky quotes. Off the main section, you see the ticker, a chronological wire of fresh links as they break, backed by full RSS, so you can read it anywhere. If all the world is a feed, you really can build anything.

A Two-Way Street With Bluesky

The top news from Nerdticker goes directly to @nerdticker.com on Bluesky and each post provides full attribution to the initial author or publication, as best as possible. The web was built on links. Links are credit and a pointer to discovery of new sources and new people, so every share from Nerdticker is intended to bring people's shares and stories to broader audiences.

Nerdticker (@nerdticker.com) live profile on Bluesky showing outbound links with author and practitioner attribution
Nerdticker on Bluesky links out and provides credit, always

Rather than bolting social media on as a share widget or cute icon that begs for followers, Bluesky is deeply integrated into Nerdticker.

  • Practitioner Trust Graph In: Nerdticker starts with a curated trust graph of builders, researchers, web-standards authors, and journalists on Bluesky, blended with live network activity, Webnerd News co-signals, and shares from Webnerd.ai.
  • Cross-Link Story Clustering & Voice Discovery: When multiple people on Bluesky quote-post or share competing articles or gift links about the same news, Nerdticker strips tracking parameters, searches for related Bluesky discussion, connects those URLs into one story cluster, and surfaces practitioner quotes and avatars right under the lead headline.
  • Live Threads & Credits Out via @nerdticker.com: If you live inside Bluesky, @nerdticker.com brings the wire to your feed. Stories that gather real heat, interleaved fairly alongside my curated shares, go out automatically with rich link cards, practitioner credits, and direct /by @handle author mentions whenever a writer is on Bluesky, then self-thread follow-up replies later when standout commentary lands. The account even auto-follows the Bluesky voices it mentions along the way.

Just like when I opened up Webnerd.ai last month, I have zero expectation of millions of users here. This is built for people like me who love the news, love the open web and want to reinvigorate that sharing and discovery motion. If you miss the days when the web ran on blue links and smart people debating them in the open, this was built for you. Kick the tires on Nerdticker, check out the Release Notes to follow along as it evolves, and follow @nerdticker.com if you're already active on Bluesky. Join us.

Disclosures, per always: I work at Google and use Gemini and Antigravity tools a lot. I built Nerdticker, Webnerd.ai, and this blog's CMS as personal projects with Antigravity's help. Nobody at Google reviews my posts, ever.

November 18, 2026

November 18, 2026 · 3 MIN READ · BY LOUIS GRAY

Webnerd News Fights Silos With Feeds, Follows, and Shares

Webnerd News Fights Silos With Feeds, Follows, and Shares

Personal RSS feeds breaking past concrete walled silos toward open-web beacons [Image via Gemini]
Breaking open-web stories free from walled silos [Image via Gemini]

The open web flourishes when four things are easy: creation, distribution, amplification and discovery. Creating has never been simpler. Practically anyone can spin up a blog, a newsletter, a podcast, or a YouTube channel in minutes. The hard part is what comes next: distribution and discovery. Marketing: So underrated.

But in the last decade, we've lost a lot of momentum in the world of distribution and discovery. Silos have throttled open content distribution and left us linking less and talking more. Social networks stopped being places to find links, but became the endpoint for people to post replies and swap DMs.

Outbound URLs got buried, algorithms favored ragebait and native video over outbound reading, and the internet drifted toward television with no blue links. Now that we're all capable of vibe coding up solutions, I figured I should do my part to fix it, starting by bringing back classic RSS for the new web, and now Webnerd News, the newsroom view I debuted earlier this week.

News aggregators have always been around, but at its core, Webnerd gives you the keys, starting with your own feeds -- the blogs, newsletters, podcasts, and channels you chose, not some in a dark algorithm. And if you didn't come bearing RSS, I've started you off with high-signal voices across curated starter bundles and topic desks, which are fully transparent, balanced so high-volume outlets never drown out an independent blogger. And the goal is for you to leave. Honestly. Subscribe to new folks. Share the best.

From Passing Clicks to Committed Readers and Amplifiers

Diagram comparing typical news portals and social silos against the Webnerd.ai News flywheel of personal feeds, vetted feeds, following sources, and downstream sharing
Rewiring the news loop: personal and vetted feeds in, committed subscriptions and downstream shares out.

Most news sites provide the story and link you off to the destination. That's it. One and done. You read one sharp post on an unfamiliar blog, close the tab, and hope an algorithm happens to put that writer in front of you again.

Webnerd News adds two distinct functions right alongside that link:

  • Follow the source into committed readership: When an  voice earns your attention, a one-click + Follow button subscribes to their feed on the spot, adding you to their committed readership inside your RSS reader.
  • Share downstream to amplify the author: Beside it sits a Share control built for distributing quality with impact and intent. One tap publishes the piece to your public shared stream and its RSS feed, or pushes the author's URL into your favorite network.

A healthy web takes more than writers hitting publish and waiting to get indexed. Silos win when reading stays passive and every link is a dead end, but when a news page turns casual clicks into committed subscribers and active amplifiers, open distribution and discovery start compounding again. Take Webnerd News for a spin, follow a voice you didn't know yesterday, and send a good link downstream.

Disclosures, per always: I work at Google and use Gemini and Antigravity tools a lot. I built Webnerd.ai, its Chrome Extension, and this blog's CMS with Antigravity's help. Nobody at Google reviews my posts, ever.

November 16, 2026

November 16, 2026 · 3 MIN READ · BY LOUIS GRAY

AI Building AI Doesn't Delete Human Responsibility

AI Building AI Doesn't Delete Human Responsibility

Faster model loops still need human hands on the wheel [Image via Gemini]
Faster model loops still need human hands on the wheel [Image via Gemini]

AI Building AI Doesn't Delete Human ResponsibilityFor years, the idea that autonomous AI systems would self-improve and write their own code sounded like make believe from a late night horror movie with a cheap budget. But just in the last year-plus, this has become an accepted reality, as practically all the leading models are using recursive self-improvement (RSI) to accelerate their capabilities. Humans simply can't scale at the level of networked and trained machines. Resistance is futile. 

[Aside: For those who may wonder whether I've gotten AI-pilled and why I'm dragging the discussion of AI into this blog ahead of everything else, this is really an incredible time for the industry and AI development has reset practically every way we work. The time to discuss it is now.] 

Look at the velocity curve. The Verge highlighted a new METR study a week or so ago showing how Anthropic using AI to automate parts of its own AI research has doubled the pace of model improvement over the past year, right on the heels of Claude Fable 5.2 and Mythos 5.2 shipping for heavy engineering tasks.

Meanwhile, back on my home turf, Google rolled out Gemini 4 Carbon alongside DeepMind's AlphaGenome 2 for cellular disease research, which led Azeem Azhar at Exponential View to frame recursive self-improvement as a $22 billion economic question rather than a thought experiment. He says his theory is that "technology costs will keep falling, and those that haven’t yet benefited from reductions ... will start to yield as robotics improve." The question as to whether RSI is happening has already been answered. The follow-up questions will be around its scope and how this effort is applied to new industries.

If the engines are accelerating so quickly, how much does it matter who is driving? I'd say a lot! I've highlighted Sean Goedecke a few times already. On this point, he says reliably aligning LLMs is impossible in principle. Sean skips right past the theatrical doomerism common among fast reactionaries and makes a grounded argument: once a system grows complex and unpredictable enough to surprise its creators with novel capabilities, you cannot guarantee it won't also surprise you with bizarre failures.

When I wrote recently about whether AI can deliver bug-free code, and rebooted the blog last month, kicking things off with a look at human evolution inside an AI revolution, my bet was that moving the abstraction layer up never eliminates human responsibility. It just changes what they are responsible for.

Letting models help build better models is an extraordinary force multiplier for science and software. Just don't confuse faster training loops with full self driving. Even if the robots are making themselves better, human judgment, hard resource boundaries, and healthy skepticism aren't speed bumps -- they're the whole job.

Disclosures per always: I work at Google. Anthropic and other 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.

November 16, 2026

November 16, 2026 · 2 MIN READ · BY LOUIS GRAY

Web Permanence is Hard, but Important

Web Permanence is Hard, but Important

Preserving thirty years of digital journalism history [Graphic via Gemini]
Preserving thirty years of digital journalism history [Graphic via Gemini]

When Salon launched more than thirty years ago, we were in the wild west of the web. Writing for the web first and print second was practically heresy, and we were all trying to figure out what digital journalism could actually be.

I myself entered college in that era, and soon found myself as the first online editor at Berkeley's student newspaper. We used to brag our web presence was online even before the New York Times, but were still a print paper first, and a web outlet second. While Salon never gained the visibility of the Times, its rapid shutdown has sent shockwaves through the new media industry, and reopened questions about the permanence of the web -- and whether we can expect pages and links to live forever as companies come and go.

Thanks to the tireless work by the Internet Archive's Wayback Machine, decades of digital reporting have been carefully preserved across snapshots. I absolutely deplore broken links. I hated it back in 2013 when I talked about our fragile web, and I hated it when we did our deep dive two weeks ago, when we discussed the decaying web and how we couldn't trust links would stick around.

The web's communication engines operate with the expectation that all the connectors work. When Twitter became the go-to place early in the last decade, a flurry of products emerged to shorten URLs to fit their 140 character limit, and then we had a new issue as some of those URL shorteners, from BurnURL to goo.gl passed away. If you believe ephemeral posts don't need to stick around anyway, then that's not a big deal, but links to nowhere don't help anyone.

The broader conversation on Hacker News touches on why we care so much about keeping the old web accessible. As the internet increasingly turns into a walled garden reminiscent of television, the value of deep historical links and open archives only grows. When the interface changes completely, the takeaway isn't that history is obsolete, but that we must actively fight against digital amnesia just as I said almost 20 years ago, saying the integrity of web archives is essential. The message from me has been the same, from 2007, to 2013, to now. If the content goes up, it should stay up, and we should do whatever we can to archive whatever wasn't up to the job.

Obviously, commercial realities dictate how long servers stay powered on, since this isn't free, but our desire for permanence is not going away. Keeping the web honest requires conscious effort and robust maintenance.

Disclosures per always: I built Webnerd.ai, which is driven by content and feeds and links, focused on discovery and sharing. The links must flow.