AIMM Session — August 21, 2025: Agents, Cold Outreach Limits, and the Boomerang Model

“Keep your focus on throwing the boomerangs — they’re designed to come back.” — Donald Kihenja

30-Second Summary

Part AI lab, part philosophy salon. The group unpacked real-world limits of GPT-5 agents, dissected why AI-powered cold outreach crashed at the C-suite level, had an honest debate about free vs. paid communities, reflected on giving without expectation (complete with Bhagavad Gita citation), and geeked out on Alex Hormozi’s $150M+ launch weekend.

1. GPT-5 Agents: Useful — But Only for the Right Job

Donald Kihenja’s clever experiment: pointed a GPT-5 agent at a freeze-dried supplement website, asked it to compare price-per-ounce across all fruit powder SKUs. It scraped the site, built a clean comparison table. First practical agent use case heard in the group.

Counterpoint: a choral director friend tested the same agent for music research — performed worse than plain ChatGPT. Lou’s synthesis: agents shine for “manual click” tasks (buying, booking, form-filling), not for nuanced knowledge retrieval.

Lou flagged the trust issue: “I’m not comfortable with anything on my computer that can control my computer.” Dirk: “I think we lost that space already a couple of months ago.”

Takeaway: before reaching for an agent, ask whether a well-crafted Perplexity query would do the job. Save the agent for tasks requiring actual clicks and navigation.

2. AI Cold Outreach at the C-Suite: Why It Fell Flat

400 emails. Zero replies. Dirk used ChatGPT to build a prompt-engineered company analysis — tested it against 5 companies he knew intimately, nailed the pain point every time. Felt like a millionaire. Sent 400 personalized AI-crafted emails to executives. Crickets.

An AI sales company out of Switzerland told him flat out: “Cold AI outreach doesn’t work for executive search.” They’d tested with best tools, best data, best prompts.

Why: senior execs are bombarded, highly triaged, have assistants filtering everything, respond to ego and white-glove treatment. They don’t click cold emails.

Lou’s nuance: AI isn’t useless — it should power the preparation and intelligence, not the mass blast. Better play: use AI to build the intel, then get personal fast. Think stealth voicemail, hand-selected content drops, executive masterminds as trust-building vehicles.

“Just because AI made it possible doesn’t mean it’s the right play for your market.”

3. Giving Without Expectation — The Business Case for Zen

Lou traced this back to his Zen teacher in San Diego: do things because you genuinely want to, not in expectation of return. Kasimir: giving with expectation of getting back is technically manipulation — when it doesn’t work, the frustration is a signal. He reframed expectation as curiosity: “Let’s see what happens.”

Donald brought the Bhagavad Gita, Chapter 2, Verse 47:

“Act, but do not be attached to the fruits of action; never let the fruit be the motive for your action, nor let your attachment be to inaction.”

Don Back: interest, enthusiasm, and always showing up — even when it doesn’t work — are the top ingredients in any relationship that eventually converts. Most people give up at the first pushback.

Hot take: this is not permission to avoid selling. “We’re in business. That’s where the rubber meets the road.”

4. Free vs. Paid Communities: Lou’s Counterintuitive Take

Bally Binning is a year into building a free biweekly community — ~30 total members, 15 regulars.

Lou’s honest take: Free communities attract tire-kickers. Free groups bring entitlement, low engagement, members simultaneously in a thousand other free groups. His preferred model: charge something, even $47/month, as your lead magnet. Even a small price triage changes who shows up. A 12–50 person paid group outperforms a 5,000-person free group on every meaningful dimension.

“If community energizes you, if it lifts you up — do it. Any marketing technique, done long enough, works. The trick is doing the one that aligns with who you are.”

5. The Hormozi Masterclass in Patience — and Scale

Hormozi ran a 7.5-hour live launch for his third book. Lou’s math:

• ~3.5 million books sold at ~$30 = ~$105M on books alone
• 85,000 new School.com members added
• High-ticket upsell at $24,000+
• Estimated $150–175M total over a few days

The group’s focus: everyone wants the launch, nobody wants the 2–3 years of free everything that preceded it. Lou: “I give him full credit. I’m not willing to pay the price he did — but kudos to him.”

Structure: first 40 minutes = massive free giveaway. Remaining 6+ hours = pitch aimed at people running serious businesses.

School.com pricing note: Hormozi opened a $9/month entry tier (vs. $99/month) with a 10% transaction fee baked in. Lou called the funnel: “Give them 3–4 months to get 50–100 members, and the 10% fee makes the $99/month plan obvious.”

Hot Takes & Resources

• “Throw boomerangs, not lead magnets.” (Donald coined it, Lou co-signed it, Don Back said he was stealing it.) A lead magnet is transactional. A boomerang is what you put out knowing it’s designed to come back — on its own timeline.
• Imposter syndrome reframe (Kasimir): it’s a signal that you’re trying to follow someone else’s path. The fix isn’t more confidence in their playbook; it’s finding your own.
• Long video workflow (Don Back): download → compress with Handbrake → transcribe with Whisper → summarize with ChatGPT.

Try This

Build one boomerang. Not a lead magnet. Not a sales sequence. One piece of content, one introduction, one resource shared with zero expectation of return.