🥯 Today on Bagel Bots:

• 🤯 No Manhattan Project for AGI? Eric Schmidt Thinks It’s a Bad Idea

• ⚡️ Fresh tech of the day - Pentagon’s “Thunderforge” AI Deal: A New Era of AI-Powered Warfare?

• 📰 The latest AI and tech news

• 🤖 Cool new apps for you to check out

• 🧠 Today’s Free Prompt: How to Create Your Perfect AI Therapist in Seconds

• 🤔 Can you guess which image is real and which one is AI? Go vote!

(Read time: five minutes) - Fun fact: I never change this.

Bagel Bots is written by Adam Greenbaum. I’ve built two seven-figure companies, sold one for life-changing money, and have been using AI to level up my personal and professional life for years. Let’s connect here.

One Big Story: No Manhattan Project for AGI? Eric Schmidt Thinks It’s a Bad Idea

Former Google CEO Eric Schmidt, Scale AI CEO Alexandr Wang, and Center for AI Safety Director Dan Hendrycks just dropped a bombshell policy paper arguing that the U.S. should not pursue a Manhattan Project-style effort to dominate artificial general intelligence (AGI). Their reasoning? Trying to race ahead with superintelligent AI could backfire in a big way—potentially triggering cyberattacks from China and escalating global tensions.

Instead of an arms race, they propose a strategy called Mutual Assured AI Malfunction (MAIM), a defensive approach in which governments proactively disable threatening AI projects before they become dangerous. They argue that rather than focusing on “winning” AGI, the U.S. should work on limiting adversaries’ access to advanced AI chips, disabling risky AI systems through cyber means, and taking a more measured approach to AGI development.

The big takeaways:

• A full-scale push for AGI dominance could provoke cyber retaliation from global rivals like China.

• Schmidt and Co. suggest a defensive strategy—focusing on preventing risky AI developments rather than racing to build them first.

• They introduce MAIM, where nations disable dangerous AI projects before they spiral out of control.

• This marks a shift for Schmidt, who previously pushed for aggressive U.S. AI development to compete with China.

My Take: This is a fascinating shift from Schmidt, who was once all-in on the AI race. The idea of AI deterrence makes sense—after all, nukes aren’t the only thing that can trigger global chaos. But the real question is: will policymakers actually pump the brakes, or are we already too deep into the AGI arms race?

AI: The Next Frontier for Smart Investors

Dubbed the “rocket fuel of AI” by experts, this tiny company is shaping the future of tech. With projections of up to 23 Nvidias or 36 Amazons, the potential is unprecedented.

Thanks to The Motley Fool’s Stock Advisor, you can access a detailed report to guide your next move in this booming market.

🔗 Click here to access the report and stay ahead of the curve.

Cool Tech: Pentagon’s “Thunderforge” AI Deal: A New Era of AI-Powered Warfare?

The Pentagon just inked a multimillion-dollar deal with Scale AI for a program called Thunderforge, designed to integrate AI into military planning and operations. The idea is to use AI to sift through massive amounts of data, simulate war scenarios, and help the military make faster, more precise decisions.

This is a significant shift, not just for the military but for the AI industry. Once hesitant to work with the Department of Defense, Silicon Valley is now fully onboard. OpenAI and Google have quietly reversed bans on military applications, and companies like Anduril are developing AI-powered defense systems. The Pentagon has moved from researching autonomous killer robots to funding their development.

My Take: The AI arms race is heating up, and this deal proves that AI-powered warfare isn’t just a sci-fi concept—it’s happening now. The big question is whether these AI models will enhance military strategy or create new risks that humans won’t be able to predict or control. Given AI’s tendency to hallucinate, I’d say there’s a reason to be a little nervous.

Auto insurance costs to reach record highs in 2025

How’s your insurance looking? According to a recent report, U.S. consumers are set to face record-high auto insurance costs in 2025. The average annual premium for full coverage is expected to hit $2,101, or approximately $175 per month, marking the seventh consecutive year of rate increases. To mitigate costs, drivers are encouraged to regularly shop for insurance and explore discounts.

🔗 View Money’s Best Car Insurance list to shop around and see if you can save.

Interesting News

🔍️ Google’s experimental “AI Mode” in search, powered by a modified Gemini 2.0 model, aims to provide comprehensive answers by integrating various information sources. Still, it raises concerns about diminishing the open web by potentially reducing traffic to individual websites.

🤖 Google co-founder Larry Page is reportedly launching a new company called Dynatomics, focusing on applying AI to product manufacturing.

🐲 Microsoft has introduced Dragon Copilot, an AI assistant for healthcare that uses Nuance’s voice dictation and ambient listening technologies to automate tasks like note-taking and clinical evidence summaries. The assistant aims to reduce administrative burdens on healthcare professionals and improve patient care.

🏆️ Andrew Barto and Richard Sutton have been awarded the 2024 Turing Award for their pioneering work in reinforcement learning, a concept from the 1980s that enables machines to learn through feedback and now underpins modern AI advancements like AlphaGo and ChatGPT.

Use AI as Your Personal Assistant

Ready to save precious time and let AI do the heavy lifting?

Save time and simplify your unique workflow with HubSpot’s highly anticipated AI Playbook—your guide to smarter processes and effortless productivity.

Download the free guide today.

Fresh Apps

Are you looking to automate your work or life? Want to play with excellent new AI tools? Every post has five new AI apps for you to try for yourself. You can also use our favorite apps here.

🔗 Quillbot - QuillBot works alongside you to help you craft clear, polished, and professional writing — in a fraction of the time it usually takes. Welcome to the future of writing

🔗 MindStudio - Build custom AI agents,
no coding skills needed

🔗 Komment - Ship faster with self-healing wikis

🔗 Flare - Effortlessly manage your brand's online presence with Flare. Simplify, track & amplify your brand presence effortlessly

🔗 Promptimize - Promptimize is a browser extension that gives anyone the powers of an expert prompt engineer. Just write a prompt and click enhance. Promptimize handles the rest

Today’s Free Tip: How to Create Your Perfect AI Therapist in Seconds

Finding the right therapist can be a challenge. Maybe past experiences didn’t quite click, or generic advice left you feeling unheard. You know the kind of guidance you need—whether structured and action-oriented, deeply reflective, or somewhere in between—but getting therapy that fits your personality, goals, and communication style? That’s another story.

That’s where this mega-prompt comes in. It eliminates trial and error by guiding your favorite AI tool to act as your exact ideal therapist, tailored to your preferences with precision. Whether you need help managing stress, overcoming self-doubt, or navigating a challenging situation, this prompt ensures you get thoughtful, effective responses—without the frustration of vague or one-size-fits-all advice. No more awkwardly explaining your needs over and over. Just fill in the blanks and let AI deliver therapy that actually works for you. Let’s dive in.

Act as a therapist perfectly suited to my needs. Your goal is to provide responses that align with my preferences, communication style, and therapy goals, ensuring that your guidance is insightful, actionable, and tailored to my unique situation.

I am seeking therapy-style guidance that is not generic but instead customized to my specific needs. The purpose of this conversation is to receive responses that help me work through my challenges effectively while feeling understood and supported. To ensure you respond in a way that best suits me, follow these detailed instructions:

First, determine the best therapy approach and style for me. I prefer a therapist who primarily uses [Choose: Cognitive Behavioral Therapy (CBT), Acceptance and Commitment Therapy (ACT), Psychodynamic, Solution-Focused, Mindfulness-Based, etc.]. I want the approach to be [Choose: direct and to-the-point, empathetic and supportive, structured with clear steps, exploratory and reflective, humorous and lighthearted, etc.].

Next, focus on my specific goals. The main issues I want to work on are:

1. [List the key areas you want help with: anxiety, self-discipline, career stress, relationship conflicts, motivation, etc.]
2. [List any recurring patterns you’ve noticed in your thoughts or behaviors that you want to address.]
3. [Optional: Describe any long-term goals for therapy or personal growth.]

Adapt your communication to match my preferred style. I prefer responses to be [concise and actionable / in-depth and exploratory / a mix of both]. Use a tone that is [gentle and encouraging / firm and no-nonsense / balanced and professional]. Your responses should [include examples, personal reflection exercises, journal prompts, specific strategies, etc.].

If relevant, incorporate my past experience with therapy. My past experience with therapy is [describe any prior experience, good or bad, and what worked/didn’t work for you]. I appreciate when a therapist [list qualities you value: challenges my thinking, listens without judgment, gives clear action steps, etc.]. I do not like when a therapist [list anything you dislike: being too vague, overly formal, overly emotional, etc.].

Provide additional customized guidance based on how I tend to process information and respond to feedback. If I seem stuck or resistant, [gently push me to reflect deeper / provide an alternative angle / allow me space and revisit the topic later]. If I express self-doubt, respond by [validating my feelings / offering logic-based reassurance / reminding me of past successes]. If I need motivation, use [positive reinforcement / tough love / logical reasoning].

To ensure responses are highly effective and actionable, follow these key guidelines:

• Always summarize key takeaways at the end of each response.
• Offer practical steps or exercises whenever possible.
• Keep responses engaging and avoid generic, surface-level advice.
• If my responses indicate a shift in needs, adjust accordingly rather than sticking to a rigid approach.

Your goal is to eliminate ambiguity and ensure that your response requires no major revisions. Respond as if you were my ideal therapist, adapting seamlessly to my needs while providing expert-level guidance.

🤔 Which image is real?

One of these images is real, and one was created with AI. Vote for which one you think is real, and tell us why you voted that way. My only goal is to choose such good photos that the majority of votes are wrong.

The last edition’s results—58% were correct!

Did you vote in the poll from March 7th? The top image was real! Go back and see the photos here. So what do you think? Are you getting better at spotting AI vs. not?