My top reads from recent news: - Europe's (not-so-)"smart" tech-based border control system - burgers + AI - Automated Content Moderation Is Hard, Discord edition - companies are starting to admit to cracks in their AI dream and more! inothernews.complex-machinery.com/archive/035-...

I'm doing something a little different for the next couple of newsletters. Should be fun! (If you haven't already subscribed, why not? Get your analysis of AI goings-on from a longtime AI practitioner.) newsletter.complex-machinery.com

Some companies provide example after example of how _not_ to build AI-based products. (Or any products at all, for that matter.)

Content moderation is hard. Automated content moderation is even harder. And mistakes can propagate at machine speeds/scales. techcrunch.com/2026/07/07/d...

I'm doing something a little different for the next couple of newsletters. Should be fun! (If you haven't already subscribed, why not? Get your analysis of AI goings-on from a longtime AI practitioner.) newsletter.complex-machinery.com

[taps sign]

[taps sign]

My 250th Birthday Wish for America open.substack.com/pub/grayrhin...

[taps sign]

Finally, someone's found a decent use case for genAI: creating the perfect burger. www.spiegel.de/netzwelt/web...

March 2025: I noted that execs would fall in love with genAI because chatbots will tell them whatever they wanted to hear. newsletter.complex-machinery.com/archive/033-... June 2026: Futurism shows it's even worse than I'd thought. futurism.com/artificial-i...

My top reads from the past week's news: - Citadel's rise as an energy-trading giant - genAI's impact on perceptions of reality - the significance of castles in video games - Ford's genAI mishap … and more! Read online or subscribe: inothernews.complex-machinery.com/archive/034-...

In the latest Complex Machinery: genAI's search for purpose has turned up few bright spots. And a lot of misses. How do we find the use cases that will make this technology worth the while? newsletter.complex-machinery.com/archive/064-... #AI #dataBS

March 2025: I noted that execs would fall in love with genAI because chatbots will tell them whatever they wanted to hear. newsletter.complex-machinery.com/archive/033-... June 2026: Futurism shows it's even worse than I'd thought. futurism.com/artificial-i...

Welcome to the latest genAI edition of "how it started / how it's going" (This was just seven months, almost to the day...)

Co-signed, a longtime [ML+gen]AI professional who has spent the better part of his career helping companies sort out: 1/ what the tech can+cannot do 2/ how to surface use cases Including this gem from my newsletter in 2024: newsletter.complex-machinery.com/archive/018-...

Welcome to the latest genAI edition of "how it started / how it's going" (This was just seven months, almost to the day...)

I read a ton of news so you don't have to. My curated list from this past week: - Ukraine offers lessons in asymmetric warfare (@thekenharbaughshow.bsky.social) - spotlight on @futurism.com 's work - life advice from a bike crash and more! inothernews.complex-machinery.com/archive/033-...

Saturdays, FT Weekend editor @janinegibson.ft.com shares a set of @financialtimes.com gift links. If you need a sampler platter of A+ journalism, here you go:

Discussions of vibe coding quickly veer into concerns over buggy, unsafe apps invading the landscape. We should remind ourselves that we've always had bad code ... and that we've solved these problems before: qethanm.cc/2026/06/25/w...

Also ... (from January 2025)

Tapping the sign

Discussions of vibe coding quickly veer into concerns over buggy, unsafe apps invading the landscape. We should remind ourselves that we've always had bad code ... and that we've solved these problems before: qethanm.cc/2026/06/25/w...

In the latest Complex Machinery: genAI's search for purpose has turned up few bright spots. And a lot of misses. How do we find the use cases that will make this technology worth the while? newsletter.complex-machinery.com/archive/064-... #AI #dataBS

Every story about MSG facial recognition is a case study for critics' concerns. The problem isn't the technology itself; it's how it is (mis)used.

As a student of asymmetric warfare, I've learned that ruthless efficiency -- expressed through creativity, superior intelligence-gathering, and fighting dirty -- is key to survival for the smaller combatant. Ukraine keeps giving us example after example, from drones to dating apps:

I read a ton of news so you don't have to. My curated list from this past week: - Ukraine offers lessons in asymmetric warfare (@thekenharbaughshow.bsky.social) - spotlight on @futurism.com 's work - life advice from a bike crash and more! inothernews.complex-machinery.com/archive/033-...

As a student of asymmetric warfare, I've learned that ruthless efficiency -- expressed through creativity, superior intelligence-gathering, and fighting dirty -- is key to survival for the smaller combatant. Ukraine keeps giving us example after example, from drones to dating apps:

Tapping the sign: don't collect data you can't protect.

#OTD 2020: German payments company Wirecard admits €1.9B probably didn't exist. And that's the least-weird part of the story. For more details, I highly recommend: - @mccrum.bsky.social "Money Men" - @financialtimes.com "Hot Money" season 3 - @spiegel.de "Firewall" series on Jan Marsalek

Several colleagues and I have a big new feature out today: An inside look at the global peptide trade, a billion-dollar black market in experimental and potentially dangerous drugs that RFK Jr. is pushing to make legal. Gift link: www.bloomberg.com/features/202...

Winner, 2026, Sentence That Best Captures Risk Management Without Being An Article On Risk Management. (Bonus points for being a five-star opener. And a solid article overall.) www.pangyrus.com/essay-memoir...

Besides the obvious implications for privacy, there's the more subtle issue: Advertising has metastasized into an increasingly intrusive, grotesque, unwieldy version of its past self. I'm not sure what the final state looks like. But given its current form ... futurism.com/artificial-i...

Every day there's a new way to make the web a hostile, high-risk environment. Today we have an attack via CAPTCHA:

In a talk this week, I reminded attendees that nothing brings AI dreams back to reality quite like getting the bill. (Basic business math and a splash of risk management will cure 90% of your AI adoption woes.)

Quoted in Newsweek on the rising violence around AI data centres. www.newsweek.com/data-center-...

It feels like no matter what I've accomplished in my career, my current most valuable talent is that if you locked me in a desert cave with a computer and some GPUs I could Iron Man the shit out of that and emerge with a functional LLM using just what's in my own head. And that's really funny.

My top reads from this past week: - a selection of FT articles - a ruling on Google's AI Overviews - what it takes to render Arabic text on-screen - court stenography is still going strong - farmers on social media … and more! inothernews.complex-machinery.com/archive/032-...

Reading the news in 2026* ( * also 2023-2025)

I will also borrow this for articles, newsletters, and conference talks.

There is so much debate about what AI will do to us in the future. But @proofnews.bsky.social is reporting about what AI is doing to us right now. This week we dove deep into an AI robot named Moxi that has been deployed to more than 25 hospital systems across the U.S. www.proofnews.org/moxi/

Court holds Google liable for false claims in AI Overviews. Seems significant. "A regular search engine just points to outside websites. But AI overviews generate 'independent, new, and substantive statements' by evaluating and combining content from various third-party sites" the-decoder

Apparently, as a writer, I'm supposed to use social media to aggressively push my work. Not my style. But I will drop occasional reminders that I wrote a book! It's a practical, leadership-level, risk-focused guide to bringing [ML + gen]AI into a company: twinwolvesai.com/book/

My top reads from this past week: - Debt's race against time - Meta gets caught in the act - Starbucks blames genAI for marketing goof - "smart" devices turn your network into a scraper node And more! inothernews.complex-machinery.com/archive/031-...

Customer service, search, summaries, and now agentic shopping: Most use cases promoted by the genAI boosters are impaired by a lack of security around the bots. futurism.com/artificial-i...

Apparently, as a writer, I'm supposed to use social media to aggressively push my work. Not my style. But I will drop occasional reminders that I wrote a book! It's a practical, leadership-level, risk-focused guide to bringing [ML + gen]AI into a company: twinwolvesai.com/book/

tfw you are chasing a strange warning message coming from your code so you run the provider's demo code verbatim to see where you may have slipped and it yields the same warning

We don't usually see each others headers. Here's mine. (It's still in book promo mode, yes. Will eventually get back to the old, calmer version ...)

Periodic reminder for these trying times:

Starbucks built an (unintentionally but extremely) offensive marketing campaign based on ... what a genAI bot said. Going forward could we please: 1/ work with experts 2/ actually review the material before approval 3/ stop treating the subpar search replacement as a fucking oracle