😺AI's impact at work

PLUS: A new DeepSeek V3!

Welcome, humans.

This week’s already been a whirlwind here at The Neuron. Point in case: Our team just had an intense debate about whether we should be covering more tutorials or news in 2025.

Instead of us guessing what you want, we figured we'd just ASK YOU directly.

We've put together a quick 3-minute survey to understand what you actually want from The Neuron. More news? More tutorials? Less jargon? More jargon? (kidding on that last one).

Take our quick reader survey here

Your feedback will directly shape what lands in your inbox this year. And y’know, what lands in the inboxes of 500K+ other AI enthusiasts, too. No pressure!

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

A new study dives into AI’s impact in the workplace vs humans working alone.
Microsoft launched six Security Copilot agents.
OpenAI's COO is taking on more responsibility from Sam.
Wired surveyed developers on AI with fascinating results.

Advertise in The Neuron here.

AI + humans > teams? New Harvard study shows GenAI can be your “cybernetic teammate”

You've heard of work wives and work husbands, but what about work AIs?

A fascinating new Harvard study suggests your AI assistant might be just as valuable as that coworker who always brings snacks to meetings (if not more so).

Researchers studied 700+ Procter & Gamble professionals to see how AI (in this case, ChatGPT 4 and 4o) affects teamwork, and the results are honestly kind of mind-blowing.

The experiment compared four scenarios: individuals working alone, teams without AI, individuals with AI, and teams with AI.

Here's what they discovered:

Teams with AI were 9.2 percentage points more likely to produce top 10% solutions—about 3x better than people working without it.
AI users finished tasks 16.4% faster while writing more (quantity and quality, the dream combo!)
People using AI reported more positive emotions than those working without.
Weirdly, AI users felt less confident despite doing objectively better work—AI imposter syndrome, am I right?

Perhaps most fascinating: AI broke down workplace silos like nothing else. Without AI, everyone stuck to what they knew—technical folks stayed technical, commercial people stayed commercial. With AI? Suddenly everyone started speaking each other's languages.

The researchers coined a term for this phenomenon: the “cybernetic teammate” (referencing this)—which sounds way cooler than “productivity tool.”

Here's what this means for you: Organizations that view AI as just another productivity tool are missing its true potential. AI researcher Ethan Mollick highlighted the following ways to apply these learnings:

Rebuild teams to break down walls between departments.
Help staff contribute valuable ideas outside their usual expertise.
Enable AI-equipped individuals to handle work that once required team-wide input.
Remember that AI can provide some of the good feelings of traditional teamwork.

And remember—the researchers say this study might actually understate AI's impact. Since participants barely knew how to prompt AI effectively, we're likely seeing the floor, not the ceiling. As people get better at using AI and models improve, these performance gaps could get much wider.

Oh, and P.S: P&G isn't just filing this study away—they're already implementing these insights across their innovation pipeline. This is a preview of AI-transformed workflows in action. As one commenter noted on Ethan’s blog post:

“We are potentially entering a productivity divide between those who utilize AI properly and those who do not…If you are rejecting AI as a matter of principle you will no longer be able to compete, even if you are the smartest person in the room.”

That's a polite way of saying: get on board or get left behind.

FROM OUR PARTNERS

Here’s why Attention is all you need…

We’re big Attention fans at The Neuron. Here’s why: Attention's platform is transforming how companies sell.

Example: HorizonIQ boosted post-call summaries from 50% to 99%, saving 25 minutes per call and doubling productivity.

Here’s how it works:

Tell Attention exactly what to pay attention to in a meeting (goals, budget, etc).
Attention listens in to the call.
Attention outputs important insights, action items, and follow-up emails.

There’s plenty more features, too. If you sell anything, literally drop what you’re doing and go book a demo with Attention ASAP.

Prompt Tip of the Day

Level up your AI responses with this interview technique:

Before proceeding with this task, please ask me any questions you need to provide the most helpful response possible. Consider aspects like context, specific requirements, format preferences, and any constraints I should be aware of.

When you add this to the end of your prompt, your AI should respond with targeted questions about details you never would’ve considered yourself, dramatically improving output quality (similar to how Deep Research asks follow up questions before diving in).

One more tip: if you’re asking for help and trying to learn, add the text below onto your prompt as well. It’ll transform overwhelming walls of text into manageable instructions you can actually follow:

Once we've clarified the details, please present your response in a step-by-step format, pausing after each step so I can process the information before moving to the next point.

This is really helpful if you’re trying to learn a new concept, like AI researcher Andrej Karpathy shared on X. He was able to build an entire iOS app in Swift (Apple’s coding language) without prior Swift experience.

As he noted, “I didn't even read any docs at all, I just opened a ChatGPT convo and followed instructions.”

He simply followed ChatGPT's step-by-step instructions and had the app running on his physical phone within about an hour.

Treats To Try.

There’s a new DeepSeek V3 that seems to be better at frontend coding and UI development tasks, and could be the foundation for the company’s new R2 thinking model (expected in April).
Hunyuan-T1 from Tencent is a new reasoning model that aims to rival DeepSeek on performance and price—try it here.
Midjourney now lets you organize your images with folders, describe web images to generate prompts, and create personalized moodboards that blend with style references (example).
SuperDial makes outbound healthcare calls to insurers for you, eliminating hold times for verifications, authorizations, and claims.
Base44 transforms your ideas into working apps, giving you everything you need built in—it’s a new launch, so stability might be spotty; plans range from free (25 messages / month) to premium ($20-$200/month).
n8n lets you build complex AI agent workflows by typing what you want instead of coding, connecting 400+ tools (raised $60M).
Bee Pioneer is a wearable device that captures your conversations and experiences, turning them into summaries, insights, and reminders you can access anytime—read more or watch this.

See our top 51 AI Tools for Business here!

Around the Horn.

Microsoft launched 6 new Security Copilot agents that manage phishing alerts, patch vulnerabilities, and prioritize threats that’ll be ready in April.
OpenAI released an update to ChatGPT’s Advanced Voice Mode.
The head of US-China relations urged for AI cooperation between the two countries despite tensions to avoid duplicating efforts.
Gemini's new premium AI features let it see your screen or camera feed in real-time.
OpenAI gave COO Brad LightCap more responsibilities to oversee OpenAI’s day to day operations while Sam focuses more on research and products.
Browser Use raised $17M to help AI agents browse the web using text descriptions vs reading code.
Wired surveyed 730 devs about AI, and found 75% have tried it, 38.4% are pessimists, 35.6% are optimists, 26% are agnostic, and 17% use it all the time—lots of eye-opening facts in this piece!

FROM OUR PARTNERS

Google Cloud’s Future of AI: Perspectives for Startups

Want unique perspectives on the most critical cloud and AI trends, opportunities, investment areas, and challenges that startup leaders need to know in 2025??

Hear from 23 industry leaders, including Amin Vahdat (Google Cloud), David Friedberg (Ohalo Genetics), Harrison Chase (LangChain), Crystal Huang (GV), and more.

Download here.