- Designspun
- Posts
- You Spin Me Right Round: AI’s Dizzying Week
You Spin Me Right Round: AI’s Dizzying Week
It was a very busy news cycle in AI this week: announcements from Google, OpenAI, and Anthropic.
The news cycle reinforces the essay of the week. AI feels overhyped, but it’s here to stay. The frontier labs are constantly releasing more capable models. Claude 4 improves on 3.7—which was already the most widely-used and consistently good coding model—regaining its top position in benchmarks. By acquiring Jony Ive’s hardware startup, OpenAI is thinking years ahead and exploring how AI-powered devices can help us in our daily lives. And Google’s Gemini is being inserted into pretty much all of Google’s ecosystem.
The AI Hype Train Has No Brakes
Two years ago, I thought AI was overhyped. I was wrong. Google Trends shows interest in AI has doubled in the last 24 months, and we’re nowhere near plateau.
But here’s what’s really wild: We’ve gone from ChatGPT being a "stochastic parrot" to AI systems that can plan, research, and write 10,000-word papers with citations—all in just two years. Former Google CEO Eric Schmidt says this could drive 30% annual productivity increases that economists literally have no models for.
The flip side? The "AI 2027" scenario suggests we could lose control of AI systems within the next few years, with some experts putting human extinction timelines around 2031.
So what’s actually happening? Is AI underhyped or are we racing toward disaster? And what can those of us building AI-enabled products actually do about it?
I dove into the latest expert opinions from Schmidt, The New York Times, and AI safety researchers to make sense of where we’re headed—and what we can control.
Sam and Jony Introduce io
Speaking of hyped, I’m sure the timing of this announcement was to drown out the news Google was expected to make because they were holding their developer conference this week. And I think OpenAI succeeded. All anyone could talk about was this partnership, how this is Apple’s death knell, and then what the actual device could be.
After a few more days of ruminating on the news, I am convinced by Sam Altman that there needs to be a better way of interfacing with AIs beyond opening a your laptop, going to a web browser, and typing into a chatbox. How can we make it even more useful all the time?
More details on the acquisition are in the video.
Highlighted Links
Junior design jobs have been harder to come by. I’ve been worried about how the next class of junior designers—and all professionals, for that matter—will learn in the age of AI. Patrick Morgan has some suggestions. |
Google’s AI Overviews are transforming search behavior, with users increasingly getting answers without clicking through to websites—a shift from traditional searching to "prompting" that could fundamentally reshape how we navigate the web. It’s a good read before diving into this week’s Google I/O news. |
What I’m Consuming
At Google I/O, everything is changing and normal and scary and chill. What did Google announce at their annual developer conference? (Casey Newton / Platformer)
Introducing Claude 4. Beyond the announcement of these new models’ capabilities, the video about how Anthropic employees use Claude everyday is fascinating. (Anthropic)
Figma CPO on How Figma Builds Products. Great interview with Yuhki Yamashita, Figma’s CPO and former Head of Product at Uber. (Harry Stebbings / 20VC)
Sam Altman, OpenAI and the Future of Artificial (General) Intelligence. Two authors of recently-released books about OpenAI discuss Altman and the company with Kara Swisher. (On with Kara Swisher)
The Roundup: AI creative workflow, vibe coding, ethics, agents, and more. Aarron Walter and Eli Woolery use a highlights episode to reflect on how AI has affected the design profession. (Design Better)