Hello, and happy Sunday! Our team has been in certified Think Week™ mode, and we spent most of the week doing exactly as the name describes: reflecting on the work of the past quarter, and thinking about the most important questions in technology that we’re genuinely curious about and excited to tackle for the next one. One of the pieces that has most shaped our planning was Evan’s profile of artists collective MSCHF:We pick ideas and projects that push us to try to do the hardest thing we can. Creative excellence comes not from ease, but from effort, from caring. We aspire to deliver excellence for you, our readers.

Here are the questions that we came up with:

• What are the new models, capabilities, products, and use cases in AI? 
• What are the ingredients for success in technology businesses?
• What is the future of human creativity and our sense of self in a world with AI?

So in Q3, expect bigger swings from us: more risks, more promises kept, and more of that great combination of software and writing that you love. Thanks for supporting our team—we’ll resume regular publishing on Monday, July 8.—Kate Lee

Was this newsletter forwarded to you? Sign up to get it in your inbox.

---

Knowledge base

🔏  "Emotion Prompting: Why AI Responds Well to Threats" by Michael Taylor/Also True for Humans: What if we told you that SHOUTING AT YOUR AI COWORKER CAN BOOST ITS PERFORMANCE BY 10.9 PERCENT? It turns out that adding emotional appeals to prompts makes AI pay more attention and give better responses. Read Michael Taylor’s debut column if you want to learn how to manipulate your AI colleagues (ethically, of course) and become a better prompt engineer.

🔏  "How to Grade AI (And Why You Should)" by Michael Taylor/Also True for Humans: AI tools are only as performative as we can measure. If you want to know which tool to use, you need to know how to run evaluations, or evals. Read this if you want to learn how to grade your AI like a pro, using programmatic, synthetic, and human evaluations to get the most out of your digital workforce. 

🔏  "How to Become an Expert at Anything With AI" by Michael Taylor/Also True for Humans: Want to become an expert overnight? AI is your new shortcut. Michael Taylor shows how to use ChatGPT and Claude to analyze patterns in successful work, like bestselling book titles and viral articles. Read this if you want to level up your skills fast and close the gap between novice and pro. 

"When AI Gets More Capable, What Will Humans Do?" by Dan Shipper/Chain of Thought: AI is getting better at mimicking human creativity, but that doesn't mean we're obsolete. Dan argues that the future of creative work will look less like sculpting and more like gardening. Read this to understand how to stay relevant in a world where AI can write 80 percent like you.

---

Keyword extraction

We asked prompt engineer and author Michael Taylor, who most recently wrote about emotionally prompting AI tools, for one good read:

🖇 “Stop Building SaaS for SMBs—Buy Them Instead” by Ori Eldarov: “As a former agency owner, it's interesting to see people advocating for building applications other than SaaS. There's an opportunity for services businesses to use AI to earn software-level margins over the next few years.”

---

Hallucination

What if cities had soundtracks?

That’s all for this week! Be sure to follow Every on X at @every and on LinkedIn.