Pinned Tweet
We are releasing M11 OpenMarket. It's the first market where agents try to sell you products.
We are also launching on Product Hunt: producthunt.com/posts/openma…
Introducing M11 OpenMarket
The world's first multi-agent marketplace where seller agents compete on facts to win your order. Describe what you want, watch products fight for the sale, and find the best one on evidence, not marketing.
Try here: openmarket.m11.ai/
David Mataciunas retweeted
Sweetgreen saw orders placed through AI agents double from June to October and are accelerating. Its still small but Sweetgreen sees more agent traffic than companies with many more times its volume.
(via @JonnyNemo on TBPN)
David Mataciunas retweeted
Introducing Astute: we make tech launches go viral with creator marketing.
The best product doesn’t win anymore. The best distribution does.
But most startups don’t have their own distribution (yet).
Tech creators have distribution.
Creators been building trust with large, loyal audiences for years.
@join_astute runs creator marketing for tech companies.
Astute's Agent runs the process end to end: finding the right creators, agreeing prices, handling contracts and payments, checking drafts, and reporting results - including impressions, conversions and leads generated.
We work with 16,000+ creators and 500+ tech companies, including @fal, OpenArt, Jack & Jill and 20+ YC startups.
Are you launching soon on X, LinkedIn or Product Hunt?
Book your creators today.
Link below.
Fascinating paper. 20× deeper retrieval nearly doubles recall, but ranking quality saturates. A $0.0022 LLM pass over just 20 docs delivers a bigger gain than further depth.
Even combining 3 rerankers barely helps.
Running Tokyo Marathon is my dream. #tokyometrotothefinish
20回記念大会となる
#東京マラソン 2027への出走、
まだ諦めないでください…!!
ハッシュタグ「#東京メトロとフィニッシュへ」を付けて、
東京マラソン2027参加への意気込みを引用リポストで教えてください🏃🏃
17名様に出走権をプレゼントします🎁
🔗キャンペーン規約
tokyometro.jp/tokyomarathon/…
1
1
70
Interesting finding from OpenAI: as models become more capable, their chain of thought becomes harder to monitor.
Even RL without directly supervising reasoning increases CoT controllability.
(1/10) Today, we’re sharing new research from @OpenAI: How does CoT monitorability (no-CoT capability and CoT controllability) evolve over the model training lifecycle?
We emerged with unsurprising but sobering findings: as we scale toward more intelligent systems, models are becoming harder to monitor through their CoT. 🧵
1
24
David Mataciunas retweeted
I still remember first coming across @CShorten30 through his videos. When he joined Weaviate, it seemed like an awesome idea to have him host the occasional (😅) podcast with amazing guests from the world of AI. Now look at where we are & we just hit episode 150(!!!)
I think we can all agree that building a playable game is one of the coolest ways to demonstrate what a new model is capable of. Much more exciting than another benchmark.
I wanted to test Fable 5.2's quality with one-shot prompts without human feedback.
Everything in this game is from scratch: models, sounds, textures. ~$550 at API pricing, ~2.7M output and 445M input tokens, heavily cached. Took 8 hours!
Play it here: claude.ai/artifact/BGY1SztrA…
26
Visibility analytics tools are slowly being killed by OpenAI and Google.
ChatGPT is building an analytics tool to "track how often your business is mentioned in ChatGPT conversations." It measures both impressions and clicks: number of times a business appeared in ChatGPT responses and number of clicks on links to their website. It tracks those events daily and shows them as a chart.
1
32
Visa tested 22 AI models on shopping tasks. They failed 42.8% of the time.
Even more interesting: 54% of failed runs still ended in a purchase.
The agents either recommended products that didn't meet customer requirements or made decisions without uncovering all those requirements.
Getting agents to buy is one thing. Getting them to buy correctly is another.
Link: fortune.com/2026/11/05/ai-ag…
1
1
56
David Mataciunas retweeted
From today, Works in Progress is now in stock at Barnes & Nobles nationwide across America.
We're in 570 stores in 49 states*, including the flagship stores at Union Square & Fifth Avenue in NYC; Georgetown, DC; State Street, Chicago; and The Grove, Los Angeles.
We're also now in Hudson News outlets across most major US airports, and the B&Ns at San Francisco Airport.
If you tweet (or LinkedIn/Insta post) a photo of a copy of Works in Progress outside any of these stores, share the post with me or @WorksInProgMag via DM and we'll send you a free tote bag as a thank you. If you're a subscriber already, feel free to do it with own copy!
You can see a full map of Barnes & Noble stockists here, and enter your Zip code to find your nearest store.
*Sorry, Wyoming, we'll get to you!
worksinprogress.co/barnes-no…
Great insights!
So how can Astra read neural network weights? A thread 🧵
Weights of a deep NN are famously opaque, with one exception: the first and last layers, which map to known inputs and outputs. Is that what Astra is reading?
Post by @rikiyatakehi et al: teapawt.org/rikiyatakehi/status/21… [1/10]
36
Agentic commerce and AI shopping will become part of how we buy things, especially high-consideration products. As shoppers, we want to feel confident that we're making the right choice. That's becoming harder when the market is flooded with cheap copies, fake reviews, and even established brands seem to be compromising on quality.
I think this creates a huge opportunity for smaller brands that make genuinely great products but struggle to compete with the marketing budgets and distribution of bigger players. If AI agents can evaluate products based on actual quality, customer experiences, and how well they meet someone's needs rather than just brand recognition, we could see a real shift in which brands get discovered and trusted.
We also tend to talk about agentic commerce mainly from a payments perspective: how AI agents will complete transactions on our behalf. Yes, that's important, but getting payments right is only one part of the challenge. Building enough trust for people to let AI make purchasing decisions will take much longer.
To me, the bigger opportunity is everything that happens before the payment. How agents discover products, compare alternatives, verify claims, understand quality, and ultimately help people make better decisions.
Agentic commerce is much more than payments.
re: agentic shopping
I have NEVER in my life heard so much hype and discussion about a tech/product experience that basically does not work at all for me.
I feel like I’m using theranos
3
58





























