Been using the M5 Ultra Mac Studio for the past week and predictably it is the most powerful computer I've ever used.
At this point, the improvements are not even targeted at me (a video creator) anymore. They're targeted at being the best machines for local AI work. There's nothing I can make Final Cut do with 256GB of unified memory that it couldn't do with 128GB
It's been fascinating to see how even the way Apple talk about the computer over the years has changed. The name studio did a lot of lifting, and they showed off 3D, video workflows, development, etc. Now the site is FULL of local AI apps and use cases.
So I'm also happy to report it's a great computer that's almost completely unnecessary for every video creator like me that's not doing 3D work. You do not need to spend the $10K+ that this Mac costs. But if you do? It absolutely rips.
Grok 4.7 is now available in Visual Studio.
This latest reasoning model from xAI builds on Grok 4.6 for agentic coding and complex, multistep workflows. Select it from the model picker in Copilot Chat to get started.
Learn more: https://t.co/ycBjwQu6Ka
#VisualStudio #GitHubCopilot #AI
I built a Photoshop-like image editor. It’s called Compositor.
https://t.co/5EMXCgaWI1
I originally built it for myself to get off my Adobe subscription, but decided to release it for free and make it open source. It has all the essential tools I need for compositing, with none of the BS.
I know this workflow is kind of archaic in 2026, but I’m still using it until AI can actually get pixel-perfect on some of the details.
The entire app is 12MB. Photoshop is 6,455MB on my machine.
I had the checkmark on my old account before it was a paid feature, so I've been reluctant to pay for it on this new account. But....things change. Curious to see what the $8/mo gets me that the free version doesn't.
Microsoft's Director of Applied Science called AI scraping the largest theft of labor in human history. His presentation found Copilot's answer engine cut NYTs click-through by as much as 93 percent. https://t.co/r7HI64mvQT
Google just released a free 2-hour course on Graph & Loop engineering: 1 prompt → 100 agents → loops → graphs, from 0% to 100%:
0% → 0:35 - graph engineering from scratch
30% → 31:17 - build your first agent graph
45% → 43:40 - run hundreds of agents in parallel
75% → 1:04:58 - loop engineering: route, check, repeat
100% → 1:30:09 - self-improving graphs that work while you sleep
taught by the people who run this in production
Prompts → Agents → Loops → Graphs
most people are still stuffing instructions into one agent and calling it a system
they built the thing one layer up and let it decide what runs at all
same model, same tokens, completely different week
watch it today
the full graph engineering guide is below, save it while it is still early ↓
@GaryMarcus It’s like asking billionaires if they think we should tax them more. What do we think they’ll say? Of course they don’t want government oversight.
Anduril's Fury drone flew in formation with F-35s at Creech in July. AF personnel operated the system, with Anduril staff on site in support. F-35 pilots still are not controlling Fury directly, but...that part is supposed to be coming. https://t.co/5pu7B6Jms3
If agents are like well-informed interns that need a lot of guidance, then I would bet the ideal number of agents for each person is 7+/-, just like the ideal span of control for human supervisors. You need enough time for each to teach and coach.
"We've found that people who manage people and understand how to communicate what they want — inputs versus outputs — they're much better at running agents," said Ivan Burazin, cofounder of Daytona. https://t.co/2Jf2CKzQBc