Manton Reece
About Photos Videos Archive 30 days 90 parks Reading Search Also on Micro.blog
  • The moat is ambition

    Anthropic’s IPO is live. John Gruber had a short post today, making two points that I want to comment on:

    (a) there is no permanent moat around LLM technology, only a temporary one at the frontier that’s fueled by an enormous bubble pumped up by Jensen Huang at Nvidia

    While I agree it’s not a “permanent” moat — what lead in technology is actually permanent? — I think this sells short the moat. A lead can compound: it attracts engineers and lets a company take bets its competitors can’t. I still believe what I blogged a few months ago. It’s possible this all falls apart, but it’s also possible we have undervalued the advances that AI will bring. I know that sounds a little crazy given the hype.

    The moat isn’t just the gap between (for example) Siri AI and GPT-6.2 Astra, which solved open math problems, or the internal Claude models that Anthropic is putting to work in health and science. The moat is ambition. The moat is playing a completely different game. Some companies want to use AI to make the world significantly better, even acknowledging the risks, and some companies want to charge more for iCloud+.

    To balance that cheap shot, I will say that the Apple Watch has been a big step forward for people’s health.

    Perhaps I’ve become too cynical on one side, and too AI-pilled on the other. Even though society has been burned by technology in the last decade, especially social media and its impact on mental health, I still want to believe that tomorrow will be better and worth striving for. My complaints about Twitter and Facebook would’ve been hollow words without also starting Micro.blog.

    (b) even if I’m wrong on the first part, this particular cult of religious freaks and weirdos isn’t fit to lead a public company

    This part of John’s post I won’t argue with. While there’s potential for Anthropic to do good, there’s also something unnerving about how they approach it. It’s a dangerous path to go down when you start attributing consciousness to AI models. When you believe only you can be trusted to create the machine god that we will all then have to put our own trust in. A company with that mindset should not accumulate too much power.

    → 3:15 PM, Nov 19
    Also on Bluesky
  • One of the most rewarding parts of blogging is having a space to work out ideas. Sometimes my posts are quick thoughts. Sometimes they are essays that I put hours into. They are never going to exactly echo what everyone else is posting about on social media. The web should be diverse, not groupthink.

    → 3:00 PM, Nov 19
    Also on Bluesky
  • On a roll getting a bunch of AI-adjacent thoughts out on my blog, some still coming up, because it’s just in the air this week. I hope it comes through in my products that I have total respect for the customers who don’t agree with me. Everything is looped into the AI opt-out I added two years ago.

    → 2:22 PM, Nov 19
    Also on Bluesky
  • Feeling the tension in technology right now between cynicism and optimism. I’m at my best when I make room for both. But going all-in with the cynics is exhausting. Articulating what we’re for is more important than what we’re against.

    → 1:15 PM, Nov 19
    Also on Bluesky
  • I was on the fence at first about Wavelength for Mac, but we’re going to release it. The simple audio editing and blog publishing is useful to do from a Mac. Here’s a screenshot.

    A podcast editing interface shows audio waveforms with a segmented recording, alongside sidebar options for Podcasts, Posts, Discover, and Listen Later.
    → 12:55 PM, Nov 19
    Also on Bluesky
  • I’m pretty good at finding new ways to procrastinate. The holdup on launching the new Micro.blog home page design is that first I need to rebuild my old animation desk.

    → 12:35 PM, Nov 19
    Also on Bluesky
  • The fall of Jensen Huang

    The news lately around Nvidia is fascinating. My feelings have drifted back and forth. Initially I could not believe that Jensen Huang would resign. Last night I also watched a couple DealBook interviews with people who actually knew him, and I’ve read a bunch of stories from employees.

    There was a lot in Ben Thompson’s post this week about Jensen’s fall. On the American Dream:

    Huang was sent to America as a nine year old who didn’t speak a lick of English, and somehow ended up at a boarding school in rural Kentucky for misbehaving youth, where he would trade help on homework for protection from bullies. From there he went to a state school in Oregon, became an engineer, before famously ending up at a Denny’s diner where he and his co-founders decided to start Nvidia.

    Over the weekend, I re-read the October 29 and 30 articles from the NYT and WSJ that set off the “Alice Huang is a spy” narrative. Knowing what we know now, the story looks weaker, except for the details on Megaspeed’s apparent hustle to skirt export control laws.

    Ben covered that too:

    Nvidia’s immediate reaction to those initial controls, which restricted exports based on interconnect speed, was to immediately release the A800 in China, which was an A100 with a third of its interconnects disabled. That set the tone for the conflict that I believe was the true cause of Huang’s downfall: Huang strenuously disagreed with the various iterations of the chip ban, and Nvidia only bothered to do the bare minimum to abide by the letter of the law, never the spirit.

    The sentiment from tech folks seems largely divided based on what filter bubbles you’re spending your online social time in. For some people, Jensen is a billionaire with a stack of recent poor choices and he deserves little empathy. For others, Jensen is an irreplaceable leader who should not have resigned until there was an investigation.

    I haven’t bought into the conspiracy theories that this was a coordinated hit job… But I think it was a hit to AI. It’s convenient but perhaps just coincidental timing for Anthropic’s IPO this week. Of all the frontier labs, Anthropic is the most against open source and the open weight models that Jensen advocated for.

    Could Jensen make a comeback? It seems unlikely, and yet this whole saga is uncharted territory, not directly comparable to other founders who have left and then returned to their companies.

    Jensen wrote a long post on X about all of this. The part about a potential return:

    Since my departure, I’ve seen an outpouring of calls from NVIDIA’s employees, investors, business partners, and customers for me to return as CEO. Many continue to reach out to me directly as the company navigates complex technology, market, competitive, and geopolitical dynamics. NVIDIA is a once-in-a-generation company. If the Board of Directors decides that my return is in the best interest of our shareholders, our employees, the company, and America, I would be honored to answer these calls and return as CEO.

    In the earnings call, board chairman Harvey Jones dismissed that idea:

    Jensen Huang voluntarily resigned his positions as president, CEO, and board member. The Nvidia board unanimously accepted his resignation and does not intend to reinstate Jensen as CEO.

    Betting market Kalshi currently has Jensen as the next CEO at 10%. It has fallen from 14% when I first started drafting thoughts for this post a few days ago. Seems about right.

    I have mixed feelings about Jensen’s fall. He made mistakes and it’s appropriate to have accountability for that. But also, I like to see founder-led companies that aren’t afraid to have strong opinions. That is often how progress is made. Interesting people make for interesting companies. (Interesting people can also bring division and chaos, but I don’t think Jensen was like that.)

    Nvidia will now become yet another boring company. They will slowly bleed talent. $4 trillion companies don’t vanish, but I think we’re clearly looking at Nvidia at a height of success that it won’t return to.

    → 10:46 AM, Nov 19
    Also on Bluesky
  • Working on syncing our books via AT Protocol, now that there’s a new convention for how to structure book metadata. Might not be noticeable right away, other than a new preference, but it’ll become more useful as more apps are built.

    → 11:22 PM, Nov 18
    Also on Bluesky
  • Working more with SwiftUI, partially against my will. 🤪 It seems to introduce quirkiness that would never happen in AppKit. Almost like the default behavior is a little off, so you have to massage it back into a Mac app.

    → 7:04 PM, Nov 18
    Also on Bluesky
  • From the Artificial trailer, the movie looks really well cast. I’m going to see this, but the Sam Altman characterization is so over the top, I think it will be purely for entertainment value.

    → 3:45 PM, Nov 18
    Also on Bluesky
Recommendations
  • Jon Hays
  • Daniel Jalkut
  • Brent Simmons
  • Vincent Ritter
  • John Gruber
  • Jean MacDonald
  • Alan Jacobs
  • Molly White
  • Ana Rodrigues
  • Om Malik
  • Adrian Vila
  • Michael Tsai
  • Dave Winer
  • Rebecca Toh
  • Ben Thompson
  • Sven Dahlstrand
  • Austin Kleon
  • Ton Zijlstra
  • Nick Heer
  • Jason Fried

Blogroll as: OPML | JSON

  • RSS
  • JSON Feed
  • Surprise me!
  • Tweets