Technology in focus
⚡ The CEO of Shopify has ruffled feathers with a controversial memo, advocating for reflexive AI usage as a baseline expectation. While we agree that using AI is the best way to understand AI, we also think organisations must be intentional about its use. As our Network Member Francine Bennett tells the Business Leader’s Voice podcast, businesses must have a clear understanding of the problem they want to solve before they adopt AI .
🔎 A long read on what went wrong for the Alan Turing Institute (ATI), according to Alex Chalmers. He reflects that “the story of the ATI is, in many ways, the story of the UK’s approach to technology”: changing priorities, too many goals, and a fatal ‘everythingism’ approach. Future work on the ATI might benefit from learnings in TIAL’s playbook on designing new institutions.
🙌 A grounded vision of AI as ‘normal technology’ from the Knight First Amendment Institute at Columbia University. Positioning AI as a tool that humans “can and should control”, this essay offers practical considerations around its impacts. Some more healthy skepticism to be found in Steven Sinofsky’s piece on the middleware problem: why Anthropic’s open source Model Context Protocol, designed to connect LLMs with other software to empower AI agents, won’t be a silver bullet. Plus some great recs in this reading list on AI and software engineering, compiled by LeadDev.
📱 Interesting read from Semafor on the role of group chats in shaping America’s rightwing elite, particularly its alliance with Silicon Valley: “The group chats aren’t always primarily a political space, but they are the single most important place in which a stunning realignment toward Donald Trump was shaped and negotiated”. Plus an interesting piece from the Verge on the politics of ChatGPT’s Ghibli filter, recently used on the White House X account, and how it functions as the aesthetic of the Trump administration.
🎨 While copyright laws don’t currently protect against AI copying art, as this piece by Creative Commons explains, it’s encouraging to see that Nvidia will embed digital watermarks directly onto AI generated content using the SynthID tool, developed by DeepMind. It’s also a much needed measure in the age of AI slop. |