1. A crystal ball for the West

    Read & react

    You're probably aware by now of the growing backlash against Flock and its automatic license plate readers, also known as ALPRs. From heated city council meetings in numerous states, to websites dedicated to tracking the weed-like spread of these surveillance tools across the country, to incidents of outright vandalism and destruction, there is a growing awareness of our shrinking ability to do almost anything without being tracked by someone or something.

    As someone living in New York City, I'm keenly aware of the difficulty in moving through the world without being tracked. When you leave your house in the city, you likely pass a dozen cameras on storefronts or residential buildings; when you reach the subway, you tap your phone to pay the fare; when you reach your destination, you're on camera in the subway station; when you get to work, you swipe a key card to enter the building that logs the time and who you are; and on it goes...

    At least there are ways around some of these tracking points. For example, you can stilly pay cash for your subway ride. But opting-out of a world that is increasingly focused on data-collection schemes is becoming more inconvenient and sometimes unavoidable all together. For example, in some cities and towns in this country, if you drive a car virtually anywhere you are being tracked by Flock and that data is accessible by police. There is no meaningful choice to avoid it either as many American cities have been built around the automobile.

    As the West fights its own battles with privacy, enter this concerning report on Moscow that provides a glimpse into what modern cities around the globe might look like in the near future:

    From the report:

    On the surface, today’s Moscow is a technological wonderland. You can use your face to pay for almost anything, like subway fares, or even to finalize the purchase of an apartment. Robots scurry across sidewalks making 15-minute grocery deliveries, and driverless taxis are set to begin commercial service this year.

    It sounds so futuristic, but underneath it is an increasingly centralized surveillance state.

    Moscow’s roughly 300,000 CCTV cameras make it one of the most heavily watched cities in the world. It is impossible to avoid a lens while simply walking through the city. The same cameras that whisk subway passengers through turnstiles are also used to monitor and flag people.
    These databases help the authorities target political dissent in addition to crime. Biometric feeds can be used to hunt down critics of the government after they hold gatherings. The days when outspoken Russians could flee to remote towns to avoid the authorities, as they did during Stalin’s purges, are largely over.

    The same tools that come with the promise of convenience and safety, are the same tools that can be quickly and easily misused to suit political ends. Here in the West, we've seen how surveillance tools in the wrong hands can be used to stalk people or how even seemingly innocuous databases have been co-opted and used for things they were originally intended for.

    One last thought

    For many, privacy is often a cost-benefit analysis with privacy seen as a cost rather than a benefit that is to be weighed against convenience or safety. I've written about these arguments before, in my column 'A Design of Subtle Coercion' and the Argument for Privacy I explore the problems with this cost-benefit framing. Nevertheless, it is the default way that people think about privacy.

    As such, one of the ways that privacy can be eroded is by making the collection of data mandatory, which is a heavy hand, or by making it so inconvenient that most, if not all, people will give in. The latter method appears to be the case in the article on Moscow's surveillance state:

    The CCTV network is linked to biometric databases. Submission of biometric data is voluntary in Russia, but it is encouraged in widespread advertisements, and citizens cannot take full advantage of digital services without complying.

    It's “voluntary,” but you're missing out on a lot if you don't comply. This sounds not unlike the experience of millions of Americans who cannot drive their car someplace without passing a camera that tracks them, or an application that someone needs to use for their work or personal life for which there is no reasonable alternative.

    If we cannot meaningfully choose to opt-out without, in effect, sacrificing our involvement in society, at what point are we bordering on a society that is based on literal and social contracts of adhesion?

  2. On Trump's increasing use of A.I.

    Read & react

    Slop is the term given to A.I. output that is senseless, absurd, and often not based in reality. Some might say all A.I. output is slop. I wouldn't go that far, but the adjectives I use above is how I would generally characterize it.

    While many people worry about A.I. replacing their jobs, particularly writers, creatives, and artists–many of whose work helped train these models–the President of the United States is embracing A.I. slop.

    Peter Baker writers this interesting article in The New York Times looking at Trump's increasing use of A.I. slop to portray himself as a grandiose warrior, and to portray the world as he would like to see it, not the way it actually is.

  3. Pot calling the kettle black

    Read & react

    Other than Hunter, who knows exactly what Hunter Biden's comeback plan is.

    Over the last few months or so, Hunter has made a splash. First, on X with a series of snappy comebacks toward the Trump family and administration. Then, with the launch of his Substack where he publishes almost daily. Finally, through a series of podcast interviews, including with members of the right-wing such as Candice Owens and Nick Fuentes. All of this has spurred conversations about whether he might be seeking a run at office.

    Now, Hunter has launched a meme coin, $LAPTOP, a reference to his infamous laptop that became the subject of an October surprise news story ahead of the 2020 election.

    Despite all of Hunter's protestations about the corruption of the Trump family, which has leveled corruption allegations of their own against Hunter and his father, here is Hunter now appearing to promote the same kind of get-rich-quick venture promoted by the likes of Donald Trump Jr. and Eric Trump. Hunter's meme coin is being framed as something of a tit-for-tat trolling of the Trumps and Hunter has said 20% of the supply is reserved for people who lost money on President Trump's meme coin. But, ultimately, this looks like any other valueless hype-up meme coin, and its first five hours bear that out.

    $LAPTOP launched this morning at a price of ~$37 and, as meme coins do, the price spiked within five minutes to $222.75 until, just hours after its launch, it lost more than 95% percent of its value and at the time of writing is sitting around ~$1.77.

    It's not clear why Hunter chose to launch a meme coin and it's strikingly off-brand for his new redemption arc as the prodigal son finally telling his side of the story while leveling anti-Trump-corruption barbs. But, regardless of his plan, $LAPTOP seems nothing other than a miscalculation.

    Not only has the price turned out that way, but public reception to it has also been poor. Crypto-scam investigator Coffeezilla has a very good video looking at the project and its reception. He also reveals that Hunter appears to have had this in the works since before his comeback began:

  4. Open LLMs aren't just for hobbyists

    Read & react

    Cancelled all LLM subscriptions today. It's been more than two weeks since I've touched a commercial LLM. Instead, qwen3.8 27b running locally has been incredible. It's not as fast as commercial models but it's still quick and capable. It's amazing how far quantized, open models have advanced.

    I'm not the only one. I open the paper today to see that corporations are making more use of them too:

    This could spell huge trouble for commercial providers and their investors. I was thinking that enterprise might be an area where the advantage (so long as it exists) of frontier models is worth the expense. But even some major companies are finding open-source models capable enough.

    If the future of AI turns out to be local and open, that could ripple through the rest of the economy. Trump Admin, unsurprisingly, is picking sides with the closed models:

    Mark J. Remillard (@mjr.news)
    If the future of AI turns out to be local and open, that could ripple through the rest of the economy. Trump Admin, unsurprisingly, is picking sides with the closed models https://www.axios.com/2026/07/20/ai-us-china-open-source-kimi

    But Nvidia entering it's support behind open models and it's purchase of Hugging Face certainly puts some weight is the open models corner. The question remains what Nvidia does with Hugging Face and whether it stays true to its open model and free-to-download platform.

  5. Bye bye commercial LLMs

    Read & react

    I cancelled all of my LLM subscriptions today. Turns out, I actually had more than one and didn't realize it.

    I'm a pretty reliable LLM user. I do not subscribe to the anti-LLM and doomerish way of thinking, but I also don't subscribe to the idea that AGI is here and that we'll be living in a Terminator hellscape in a matter of years.

    I use LLMs as a tool, and I think most be do (or, at least they should). I use it for research, queries, and organization. I use for managing tasks. I use it for coding and helping with my various servers and self-hosted systems.

    But recently something significant changed: It's been more than two weeks since I've touch a commercial LLM. Instead, I've been running Qwen3.8 27B locally on my server and it has been absolutely incredible.

    It's helped me design this very site, it's helped me with research into articles I'm writing, and it's helped me code and put together various projects. And it's done all of that right from my own computer and without a peep of my data going to any third-party. To me, Qwen3.8 27B has been a game changer when coupled with OpenCode, which is a terminal -based interface similar to Claude Code.

    I've been using Claude Code for more than a year to do these projects, but I always have to be careful about setting guardrails and making sure it doesn't read certain sensitive files. Too often, it does so anyway. That means a headache for me to then go an change API keys or credentials that were viewed by Claude and shouldn't have been. Now, with a local LLM, I still try to avoid such occurrences, but it's not the end of the world because that data doesn't leave my own hardware.

    Or, we can also consider the data itself. Reviewing and summarizing emails? Reading my calendar? Summarizing my notes? Depending on the circumstances, these might be things I'm not keen on sharing with Anthropic or OpenAI or anyone. Now, Qwen3.8 can review those without a worry.

    It's really been remarkable how capable Qwen3.8 is. It's not as fast as Claude Opus or other frontier models. But, given what I've seen from Qwen3.8, which is running on my capable but consumer-level hardware, I'm struggle to conceptualize what consumers and maybe even pro-sumers cannot do on something like Qwen3.8. Sure, I can conceive of a cutting-edge commercial application for OpenAI's Atlas or Claude Fable, but I think most consumers would be perfectly happy with something like Qwen3.8 and a little bit of patience.

    I've been looking forward to no longer paying for tokens, and I feel like that time has finally arrived.

Subscribe to the notes RSS feed