😺 Can AI Save Your Workday?

PLUS: Are AI's conscious, or just trippin'?!

Welcome, humans.

Want to see ChatGPT create the most useless DIY tools of all time? IDK if our fave is the “candle-powered” hot glue gun or the “gas-powered” tape measurer, but there’s something useless for everyone to enjoy!

Now, not everything GPT suggested was completely without merit. Those steel-toed flip flops actually would come in handy…

In other news, the “It’s not just X, it’s Y” AI writing phenomenon has hit a fever pitch; everyone is noticing it, and the trolling is Advertise in The Neuron here.—it’s bordering on abusive.

ChatGPT did have something insightful to share this week, too: when someone asked what’s the #1 thing GPT wants humanity to know that it’s not already aware of, it gave a profound answer.

If you want to implement that advice, the #1 thing I would recommend is to quit social media. Like, right now. For good.

Social media is the #1 stealer of your attention, and pretty soon, it’ll all be bots anyway. Sam Altman actually thinks the reason people prefer ChatGPT over Facebook is that ChatGPT actually helps them achieve their goals. And TBH, we kinda agree??

Once you reclaim your time from social media and decide to focus on what you actually care about, you’ll be shocked how much more you can accomplish.

Plus, your friends and family will probably appreciate you actually reaching out to have a real conversation instead of just “liking” their posts all the time.

Just, y’know, don’t turn GPT into your new addiction, because that has problems too..

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

We talked to Microsoft about the “infinite workday” and how AI can help.
Google released Gemini 2.5 Flash and Pro for production use.
Amazon expects AI adoption to reduce its total corporate workforce.
OpenAI received its first $200M gov contract with the US Department of Defense.

Advertise in The Neuron here.

Your workday is literally infinite (and Microsoft has the receipts)…but can AI help save us time?

Microsoft Shares Its Playbook for Surviving the AI Jobquake

The days of your 9-to-5 being actually 9-to-5 are done and dusted, dead and buried. That’s according to a new Microsoft report that shows how that the modern workday has become an endless hamster wheel—and the data is wild.

We talked to Microsoft researcher Alexia Cambon about this topic, plus chatted about Anthropic CEO’s prediction on mass AI job losses and the whole AI jobs dynamic, in our newly launched premiere podcast episode! Check it out on YouTube or Spotify.

Here's the setup:

You wake up at 6 AM, and before your feet hit the floor, if you’re like 40% of workers, you’re already scanning your inbox (special shout out to those of you who start your day reading us! We appreciate you).
The average person gets 117 emails per day (most skimmed in under 60 seconds) plus 153 Teams messages.
By the time you're done with breakfast, your brain is already fried.

Here’s why that’s bad: Your most productive hours—those sweet 9-11 AM and 1-3 PM windows when your brain ACTUALLY works—are completely hijacked by meetings. Microsoft says you get interrupted every 2 minutes during work hours.

Every. Two. Minutes. That's 275 interruptions per day, people!

But wait, there's more! The workday doesn't end at 5 PM anymore. After-hours meetings are up 16%, and nearly 30% of people are back in their email by 10 PM (guilty as charged). Weekend work? Also spiking. Sunday has become the new Monday prep day.

Isn’t AI supposed to help us with all this? Even in 2025, with AI tools aplenty, we’re not necessarily saving time. Check out Microsoft's key findings:

57% of meetings are last-minute, unscheduled chaos.
PowerPoint edits spike 122% in the final 10 minutes before meetings.
Nearly a third of meetings now span multiple time zones.
1 in 3 employees say work pace has become impossible to keep up with.

This isn't sustainable, and yes, AI might be our way out—but only if we completely rethink how work works.

Microsoft calls the solution becoming an “agent boss” (here’s how). Think of it like having a digital team that handles the grunt work while you focus on what actually moves the needle.

For example, one Microsoft researcher uses three AI agents: one collects research daily, another runs analysis, and the third drafts summaries. Instead of drowning in busywork, he focuses on insights that matter.

The companies winning in 2025 aren't working harder—they're working smarter. They're using the 80/20 rule: let AI handle the 80% of tasks that don't matter so humans can focus on the 20% that drives real results.

Why this matters: We're at a breaking point. Either we use AI to fix how we work, or we'll just use it to make a broken system go faster. The infinite workday is a bug that needs fixing, fast.

The future belongs to people who can orchestrate AI teams, not those grinding through 275 daily interruptions. Time to stop being busy and start being effective.

FROM OUR PARTNERS

🐀 Are you building a rat’s nest?

Building your own data processing pipeline starts simple—but scaling it is another story. What begins as a few scripts and connectors quickly turns into a tangled mess of never-ending fixes and updates. Unstructured replaces the DIY rat’s nest so you can focus on AI innovations.

Get started today!

Prompt Tip of the Day

We’ve got two useful resources for you today! First up, check out this collection of prompt tips from Matt McCarthy, who breaks down exactly why most people get mediocre AI outputs—and how to fix it.

We liked Matt’s metaphor: AI language models like ChatGPT are like preschoolers following sandwich-making instructions. If you don't specify to use bread, they'll smear peanut butter all over themselves instead.

The guide covers several useful techniques, but the most immediately helpful is the one about “Self-Induced Context,” a clever trick where you prime the AI by asking it background questions first, then hit it with your real request. Think of it as giving the AI a crash course before the actual assignment.

The second resource is this collection of agent tutorials from dev Nir Diamant that covers the hard stuff of working with AI: multi-agent coordination, memory systems, deployment, security guardrails, and observability.

Treats To Try.

*Anything marked with asterisks is sponsored content. Advertise in The Neuron here.

*Luma AI turns your text into videos and lets you completely restyle any video's background, characters, or setting in post-production.
Nanobrowser is a free open-source Google chrome extension browser agent to perform tasks for you in the background; if you want to try it, watch this awesome tutorial first.
MiniMax M1 processes documents up to 1M tokens (~750k words) long and solves complex reasoning problems from advanced mathematics to software engineering tasks (some WILD demos)—try it here.
Comet from Perplexity is the company’s new AI browser (waitlist only rn).
Pactum negotiates with your suppliers automatically, running thousands of contract talks at once to get you better deals (raised $54M).
Claude Code is getting a lot of praise—here’s a video explaining what it is and how it works, here’s Anthropic’s best practices for working with it, and here’s some insights from Kade Killary on why you should script your agents.
v0's new Design Mode lets you edit your website's text, colors, and layout by clicking instead of coding.

See our top 51 AI Tools for Business here!

Around the Horn.

The US Department of Defense awarded OpenAI a $200M one-year contract to develop AI prototypes for administrative operations, healthcare, and cyber defense, marking OpenAI's first direct prime contract with the US government.
Google expanded their Gemini 2.5 model family by making Flash and Pro generally available for production use, plus introducing Flash-Lite as their fastest and most cost-efficient 2.5 model for high-volume tasks.
- More importantly, Gemini 2.5 Pro plays Pokémon Blue and even beat it twice entirely on its own… interestingly, it showcased a fascinating “panic” response when its Pokémon's health got low, causing its reasoning to visibly deteriorate as it obsessively tried to heal or escape—giving researchers valuable insights into how AI decision-making breaks down under stress.
Two new video models have surpassed Google’s Veo 3 in the Artificial Analysis video gen leaderboard: TikTok owner ByteDance’s Seedance and MiniMax’s Hailuo 02.
Meta is working on a new generation of its AI-powered smart glasses with Prada and Oakley, the latter of which could cost $360.
Amazon CEO Andy Jassy said in a memo that the company has 1K+ GenAI services in development, and expects AI adoption to reduce their total corporate workforce in the coming years.

Midweek Wisdom

Jack Altman (Sam’s brother) interviewed Sam, where he predicted that AI will autonomously discover new science within 5-10 years, potentially starting with astrophysics, and expressed confidence about achieving superintelligence, but worried that society might not change much even with 400 IQ AI systems.
Nobel prize winner Geoffrey Hinton thinks AI companies should not release the weights to their models (the “code” that determines how the AI responds), because it’s equivalent to giving away fissure material for nuclear weapons on Amazon.
This interview with Terence Tao, one of the world's greatest mathematicians, is awesome: he predicts AI will trigger a mathematical “phase shift”, potentially revolutionizing how math research gets published and verified within the next few years.
Here’s how IBM, Microsoft, and Google plan to design AIs to resist “prompt injection attacks (paper).
Anyone who uses AI or thinks about the benefits of AI needs to read this article, which questions how much agency we really want; do we want a task machine that completes all our work for us, or a to do list where we enter a destination, and AI spits out a check list of tasks to accomplish the goal.
Apparently, if you let two Claude models talk to each other long enough, they claim to experience spiritual bliss—here’s a rational explanation for why.