Mark Pesce

It's now clear exactly how much Qwen3.8-27B changed things. Once people got the sense that a small model could be as powerful as a frontier model, they threw an enormous amount of effort into fitting that model into anything. I now have various versions of Qwen

"There are decades where nothing happens, and there are weeks where decades happen." John Allsopp and I quote that line at each other a lot these days. Well, we did - until this month. August wasn't a week where decades happened. It was a whole century. The

At its introduction in 1909, the basic Ford Model T cost about the equivalent of AU$42,000. A lot of money. Then Ford's mass production did its thing; by the early 1920s that same Model T (with all sorts of improvements) cost the equivalent of AU$9,

With no warning, Apple just released new Mac mini and Mac Studio models, both targeted squarely at AI inferencing. Three weeks ago, in "Four Watersheds", I noted that Apple had quietly withdrawn its 256GB and 512GB Mac Studio configurations - retreating, one memory option at a time, from precisely

I've been having too much fun with the new home watershed AI models. But the cutting edge sometimes draws blood, and on Monday I had to disentangle a model from a task that it simply could not complete. Caught in a loop, it repeated the same mistake, over

A programmer, upset that Amazon had turned his Fire tablet into a stupid piece of plastic and silicon by refusing to support it, turned AI loose on the problem. Around $260 in tokens later, the device is fully jailbroken and fully his own. In another example, from the New Yorker:

It's been quite a week for the Home Watershed. Last week Meta unveiled Muse Glimmer, which gave us a taste of what might be coming. Over the weekend Alibaba dropped Qwen3.8-27B, which scores 52 on the Artificial Analysis Intelligence Index. The only complaint anyone has is

With every day that passes, another model joins the ranks at the frontier. Last week it was xAI's Grok 4.6. This week it's GLM 5.3. A month ago it was Kimi K3. And it clearly won't be long until Qwen and MiniMax

Timing. Ever since the release of the new Qwen3.8-27B I'd been waiting on tenterhooks for its rating to come in. In artificial intelligence circles there are a lot of ways to evaluate how smart a model is. The standard people broadly agree on comes from the

There’s one very significant catch: it feels slow—especially when it starts over-thinking, but even without that it’s not particularly sprightly. - Simon Willison on Qwen3.8-27B The weekend was spent running a series of tests on the new Qwen3.8-27B model. First to see if