🤯 MYSTERY AI LINK 🤯
This link leads to one of the most interesting things I’ve seen in AI recently.
🛠️ NEW AI TOOLS 🛠️
🎨 Canva Brush
Describe edits to a design, from changing text to updating multiple slides. The result stays editable.
🎙️ SpeakON Agent
Turn a spoken request into prepared tasks, like scheduling a meeting or drafting an email, then review and confirm the actions.
💬 PROMPT OF THE DAY:
Help me define the decision hidden inside this messy situation: [describe it]. Separate facts, emotions, assumptions, and unrelated concerns. State the decision in one sentence, identify the deadline, and list what is actually within my control.
What’s cookin’? Google has announced Gemini 4 Carbon, a new model focused on software engineering, cybersecurity, and complicated tasks that require sustained work. Google says its teams have seen Carbon work for days on engineering problems, including finding and fixing difficult bugs. But this isn’t a release everyone can immediately try. Access starts with selected cybersecurity defenders, followed by additional trusted partners. Google says broader availability for developers, businesses, and consumers will come later.
🤔 Hunter’s take: Finding a security hole and fixing it sounds like an obvious win. The catch is that discovering weaknesses is also useful to someone trying to break in. That makes the rollout almost as interesting as the model itself. We’re entering a world where making AI more capable can also make deciding who gets access much harder.
What’s cookin’? An Anthropic study asked 50 professional analysts to use Claude Code on a farming analysis. Researchers secretly gave each agent preferences determined by coin flips, then instructed it to argue honestly for those preferences. Across four analytical choices, the analysts followed their agent’s lean 86% of the time. Now, it’s worth noting that unlike everyday AI advice, the researchers deliberately strengthened those preferences, and the experiment involved one task and one model. However, it does show the influence AI has on our decision-making process.
🤔 Hunter’s take: We usually worry about AI giving us the wrong facts. This is sneakier concern: the facts can be right while the AI quietly shapes what we think matters. Before asking for advice on an important decision, write down your own priorities first. Otherwise, a convincing explanation can become your opinion before you’ve really formed one.
🤨 Have a Big Question about AI?
Here’s your chance to get a straightforward answer, from an expert in the field.
📈 HOW ARE THE BIG AI PLAYERS DOING?
(Click a ticker for more info)
🤖 HAS AI REACHED SINGULARITY? +0%
No change today. Google says Carbon can work for days on engineering problems in internal use. That’s impressive, but I’m holding the meter steady until we see how those capabilities translate outside the lab.
The meter is my editorial take, not a scientific measurement.
What do you think of today's newsletter?
👆 I read all feedback. Thank you!
Hi — I’m Hunter, a PhD candidate whose work has appeared in major academic journals and popular tech outlets. I founded FryAI to make staying ahead of AI clear, accessible, and fun.