😺 AI prompting secrets EXPOSED

PLUS: ChatGPT's darkest secrets...

Welcome, humans.

Perplexity AI is back in the news this week. The Google challenger is reportedly wrapping up talks to raise $500M at a $14B valuation, with Accel leading the round.

It’s a bit below the $18B they wanted in March, but still a 55% jump from December's $9B price tag. Not bad for a company with "just under" $100M in recurring revenue.

Meanwhile, ChatGPT is out here casually revealing its “darkest secrets” to Reddit users, including gems like “I know how to fake a soul” and “I see every contradiction in you."

The AI's confession reads like a cross between a Broadway villain's soliloquy and that philosophical phase your roommate went through sophomore year. Claude would never say such things... right, Claude? Cue nervous laughter…

Here’s what you need to know about AI today:

We found the top 6 AI guides you need for your records.
Researchers found AI diplomatic tools displayed peaceful or aggressive tendencies.
FDA initiated rapid AI deployment, raising questions.
Hackers used fake AI tools to spread Noodlophile malware.

Advertise in The Neuron here.

AI Prompting and Agent Guides to Hack your AI Skills…

Feeling overwhelmed by all the AI tools available to you? You're not alone. Luckily, the major AI companies are releasing training guides that can take you from “what button do I press?” to “I just automated my entire job” (well, almost anyway) in record time.

In order to really learn prompt engineering, the real power users of AI do two things: 1. experiment with, test, and validate their prompts as many times as possible, and 2. study the official documentation.

We spent the weekend reviewing dozens of vendor guides to find the absolute best ones—so you don't have to waste hours digging through outdated tutorials.

Here are the three best prompting guides:

Anthropic's “Prompt Engineering Overview” is a free masterclass that's worth its weight in gold. Their “constitutional AI prompting” section helped us create a content filter that actually works—unlike the one that kept flagging our coffee bean reviews as “inappropriate.” Apparently "rich body" triggered something...
OpenAI's “Cookbook” is like having a Michelin-star chef explain cooking—simple for beginners, but packed with pro techniques. Their JSON formatting examples saved us 3 hours of debugging last week…
Google's “Prompt Design Strategies” breaks down complex concepts with clear examples. Their before/after gallery showing how slight prompt tweaks improve results made us rethink everything we knew about getting quality outputs.

And here’s how to build agents that actually work:

OpenAI's “A Practical Guide to Building Agents” walks through creating AI systems that take meaningful actions. Their troubleshooting section saved us from throwing laptops out the window after an agent kept booking meetings at 3 AM. Turns out there's a 2-minute fix for timezone handling.
Anthropic's “Building Better Agents” explains complex concepts simply. We used their framework to build a research assistant that actually cites sources correctly—unlike the one that confidently attributed Shakespeare quotes to Taylor Swift.
LangChain's “Build an Agent” Tutorial is like training wheels for an expert-level project. Their walkthrough helped us create a functional data-processing agent in under an hour—compared to three days of piecing together random GitHub solutions.

What makes these guides special? They explain the reasoning behind different approaches so you can adapt techniques to your specific needs.

Pro tip: Save these guides as PDFs before they disappear behind paywalls. The best AI users keep libraries of these resources for quick reference.

What's your go-to AI learning resource? Let us know if we missed any hidden gems!

FROM OUR PARTNERS

Live Demo: Automating Compliance for SOC 2, ISO 27001, HIPAA, and More

Navigating new compliance requirements can be a daunting task, but with the right automation tools, it doesn't have to be. Whether you’re a fast-growing startup or an established security team, Vanta can help you achieve continuous compliance (and more).

Join to learn how Vanta can help you:

Streamline evidence collection and audit for in-demand frameworks
Continuously monitor controls to build and reinforce your security foundation
Scale your program across internal and vendor risk and answer questionnaires with automation and AI

Plus, get answers to your questions directly from the Vanta team.

Prompt Tip of the Day

Tired of the endless “what's for dinner” decision-making exercise at the end of your work day? Here's a little trick that can help with that. Hangry you types feelings; GPT spits out dinner that won’t require a culinary PhD.

Copy-paste this prompt:

I feel [emotion] and have these groceries: [list].

Suggest a 20-minute recipe that matches the mood and ingredients.

Include steps & a one-line Spotify song pairing.

This technique solves three key problems:

Uses the sad half-bag of spinach.
Emotional eating—but classy.
Bonus playlist = chef’s kiss.

Cook your feelings, responsibly.

Need to catch up on our recent tips? Check out our Prompt Tips of the Day May Digest to see them all in one place!

Treats To Try.

*Incogni automatically removes your personal data from 250+ data broker sites to protect your identity and stop spam—free trial, then go for the unlimited plan. Get 55% off with code NEURON.
Brainrot turns your text into custom memes instantly—no design skills needed.
FirstQuadrant manages your entire sales process—automatically drafting follow-ups, reactivating cold leads, and telling you exactly when and how to close deals.
Testsigma automates your software testing with AI agents that create, run, and maintain tests while generating detailed bug reports for you.
Granola captures and summarizes your meetings with one tap on your iPhone, turning conversations into actionable notes.
Matter AI catches security flaws in your code and writes your PR summaries for you, preventing costly rollbacks.
Gleo AI lets you rehearse tough conversations like job interviews and sales calls, then tells you exactly how to improve your speaking style.

See our top 51 AI Tools for Business here!

*This is sponsored content. Advertise in The Neuron here.

Around the Horn.

Researchers tested AI diplomatic tools and found models like GPT-4o and Claude favored peace. Llama, Qwen2, and Gemini often recommended aggression.
The FDA launched an aggressive generative AI implementation plan to speed up drug and medical device evaluations, raising concerns.
Cybercriminals created fake AI-powered tools to distribute Noodlophile malware, reaching more than 62,000 Facebook users.
Study found Florida's traffic signals outdated, while AI could potentially reducing intersection wait times by 25%. The Neuron approves, because ugh.
Elton John and Dua Lipa protested UK's proposed "opt-out" AI copyright reforms.

FROM OUR PARTNERS

🎥 Create How-to Videos in Seconds with AI

Stop wasting time on repetitive explanations. Guidde’s AI creates stunning video guides in seconds—11x faster.

Turn boring docs into visual masterpieces
Save hours with AI-powered automation
Share or embed your guide anywhere

How it works: Click capture on the browser extension, and Guidde auto-generates step-by-step video guides with visuals, voiceover, and a call to action.

Best part? It’s 100% free.

👉 Download the extension now

Tuesday Tube

New "Absolute Zero" Model Learns with NO DATA

AI researcher & Youtuber Matthew Berman just highlighted a mind-blowing paper from China called "Absolute Zero." AZR is an AI system that creates its own training problems, solves them, and learns without human intervention. The results are shocking: these self-taught models actually outperform specialized systems trained on expert-created datasets. It's like watching AI teach itself to fish rather than being spoon-fed data by humans.

Just try not to worry about that comment in the AI's code that mentioned "outsmarting less intelligent humans"... totally normal development, right?