5 min read

How to Be More Ambitious With AI (From A Claude Engineer)

Claude’s own engineer says you’re asking it for too little. Three bigger asks: the real outcome, expert-grade work, fields you’d never hire.
How to Be More Ambitious With AI (From A Claude Engineer)

Top Insights

Claude Engineer Wants You to Ask It for Bigger Things

Anthropic’s Felix Rieseberg leads engineering for Claude.ai, Cowork, and Claude Code Desktop, and says most people ask AI for its smallest trick, the usual task done faster. Rieseberg’s blog tutorial shows three bigger asks.

💡 Key Insights:

  • Claude does bigger work when you name the outcome instead of the task...Rieseberg stopped asking it to fix one crash at a time and told it to keep the apps crash-free.
  • Telling Claude to act like an expert only changes its vocabulary...it keeps the markdown tables, so insist on the databases and statistical models a real data scientist would use, and never let it grade itself.
  • Experts you’d never hire can work on your quarterly deck...Rieseberg adds a journalist, a short seller, and a forensic accountant, each doing the research they’d really do, since asking Claude to write like them only produces role play.

Read the Full Post

The Marketer Behind Loop Earplugs Says AI Can Double Your Ad Hit Rate

Alexander Lodeweyckx scaled Loop Earplugs from 30 to 200 million and now runs Foresight IQ, an AI tool that sorts your ads and picks which to scale. On the Affiliate Grand Slam podcast, Lodeweyckx shows how AI lets you make 100 ads instead of 10, and where the saved time goes.

💡 Key Insights:

  • Your hit rate can climb from 20% to 40%...briefing hours now go into backend data and AI test audiences, so you know which ideas deserve budget.
  • Your judgment stays in charge of spend...Loop hit 100 million on a spreadsheet of each channel’s real lift, and it kept steering budget alongside Triple Whale and Northbeam.
  • Write down why your ads work before scaling...one brand rode 150 to 250 ads a month toward 100 million, and the next leap holds once the playbook leaves one strategist’s head.

Watch/Listen to the Full Podcast


🚀 WATCH: How These AI Copy Bots Are Producing World-Class Sales Copy 50X Faster Than Even The "BEST" Copywriters On The Market…

(Plus… They Don't Get Sick, Miss Deadlines, Or Ask For Raises Either!)

Watch the full AI Copywriting Tell-All Video Here


Prompt of the Week

This Week's Feature: The Expert Unlock

Claude knows how the best person in your field would tackle the task you just gave it. It won't work that way unless you ask, and you can't ask for methods you've never heard of.Felix Rieseberg, an engineering lead on Claude at Anthropic, wrote a prompt that closes that gap.

Drop it in ahead of any task and Claude tells you what a top professional would do that you didn't ask for:

The raw material they'd start from, the tools they'd use, and the outside check that would prove the result right. Then it rewrites your prompt to ask for exactly that.

FULL PROMPT

I want a deeper version of this task, not a bigger one. Before rewriting it: (1) say what a top professional in this field would do here that I’m not asking for; (2) name the rawest material they would start from instead of what has already been digested for me; (3) name the specific methods and tools they would use; (4) say what, other than you or me, would show the result is right. Then rewrite my prompt to ask for exactly that.

Tool to Try

Superdraft watches the world around your ecommerce brand... a heatwave, a restock, a viral TikTok sound, a marathon in your city... and turns each moment into an on-brand email draft with the reason to send attached. It learns your voice and products up front, hands your team a draft to review and edit.

AI Tool Highlights:

📡 It Spots the Reason to Send: The Signal Engine tracks weather, search spikes, local events, cultural moments, and your own stock levels, then filters each one through your products and audience... so an NYC heatwave becomes a summer-styles email while demand is still climbing.

🕵️ An Intern That Browses for You: An AI agent with its own browser hunts the live web for openings around your brand and returns a campaign brief with the proof attached... down to the Google Trends screenshot.

✍️ Drafts Arrive Ready to Edit: Compare subject lines, refine the copy, and pick the shape... Short Text, Long Text, Image Only, or Product Grid.

🎨 Your Brand, Set Once: Superdraft builds an editable profile from your website, socials, products, and goals, and you tune the colors, CTA style, and logo size yourself... it keeps learning from your notes until the drafts read like your team wrote them.

📅 A Calendar That Writes Back: Holidays, niche dates, and seasonal hooks land on the planner with a draft idea attached.

🔌 You Keep the Send Button: Superdraft stops at the draft. Review, approve, export the HTML, and send from the platform you already use... plans start at $79 a month for 100 drafts.

Try Superdraft Now

One More Need-to-Know News Story


Kimi K3.1 Pro: Half the Price of Claude Opus, Half the Made-Up Facts
Three months after Kimi K3 muscled into the top tier, Moonshot AI is back with Kimi K3.1 Pro, and this time its own scorecard puts the model level with Claude Opus 5.5 and GPT-6 Astra.

Moonshot aimed the retraining at jobs a business actually hands an AI, like digesting piles of financial documents and getting its facts straight.

📋 The Details: Start with the bill, which went the right direction. K3.1 Pro costs $2 per million tokens in and $10 out through the API, a third under K3’s rate and about half what Claude Opus 5.5 charges. It’s live today on Kimi.ai and the API, and the full weights follow by Nov 4.

🎯 Why You Need to Know: Moonshot went straight after the model’s habit of inventing facts, and the numbers moved. On its hardest accuracy test, K3.1 Pro made up 8.3% of its claims, down from 23.4% for K3.

It also set the record on a test that feeds it scanned financial PDFs with no readable text and asks for math across several of them... roughly what quarter-end looks like on your desk.

⚡ Your Move: Scan a quarter of invoices or bank statements, drop the PDFs into Kimi.ai, and ask the question you’d normally pay your bookkeeper to answer. Then check its arithmetic against the originals before trusting the next batch.

Full Story



Mind Fodder


Thanks for reading.

Until next time!

The AI Marketers

P.S. Help shape the future of this newsletter – take a short 2-minute survey so we can deliver even better AI marketing insights, prompts, and tools.

[Take Survey Here]