The Humane Pin is Dead

The $499 wearable no one wanted is gone.

Author

Nicolás Cerdeira
February 20, 2025

Hey — It’s Nico.

Welcome to another Failory edition. This issue takes 5 minutes to read.

If you only have one, here are the 5 most important things:

Let’s get into it.

This Week In Startups

🔗 Resources

An in-depth analysis of all YC’s startups.

How to nail founder-led sales.

📰 News

OpenAI tries to “uncensor” ChatGPT.

Safe Superintelligence is close to raising $1B.

Perplexity now has a Deep Research feature.

Apple is building a humanoid robot.

💸 Fundraising

Hightouch raises $80M for marketing tools powered by AI

Baseten, a AI cloud infrastructure startup, raised $75M after the release of DeepSeek.

Abridge, a startup that uses AI to build medical documents, raised $250M.

Fail(St)ory

Humane Gave Up

This week, Humane, the hardware startup behind the infamous AI Pin, announced it is shutting down.

Most of its assets have been offloaded to HP for $116 million, and sales of the $499 Pin—which no one liked anyway—have been discontinued.

What Was Humane:

Humane wanted to kill the smartphone. Their weapon of choice? A $499 wearable brooch that beamed a laser onto your hand and asked you to talk to it like you were summoning a holographic sidekick.

The AI Pin was supposed to handle calls, answer questions, translate languages, and generally make you feel like you were living in a sci-fi movie.

Humane’s pitch was bold: screens were a problem, and their AI-powered wearable was the solution. Instead of pulling out your phone, you could just tap the brooch and speak naturally to get things done. It promised to free people from their screens and create a more fluid, seamless way to interact with technology.

In reality, it mostly made users feel self-conscious talking to their chest while squinting at a faint laser display. The device’s voice recognition struggled in noisy environments, and the laser projection was almost useless in bright light.

Despite this, the idea, plus Humane’s ex-Apple founders, helped them raise over $230 million. Investors love a big vision—until it flops.

The Numbers:

  • 📅 Founded in 2017.

  • 💰 Raised over $230 million.

  • 💥 Launched the AI Pin in April 2024 for $499.

  • 📉 Sold to HP for $116 million in 2025 after the product bombed.

Reasons for Failure: 

  • Product Was a Mess: The Pin was supposed to replace your phone. Instead, it was slow, misunderstood commands, and had a laser display that was barely visible outdoors. Marques Brownlee, a top tech reviewer, famously called it “the worst product I ever reviewed.” That’s not the kind of quote you frame in the office.

  • Fire Hazard Bonus: Just when it seemed things couldn’t get worse, Humane told users to stop using the charging case because it might catch fire. Nothing says cutting-edge tech like a burning battery.

  • No Product Market Fit: Humane was solving a problem that didn’t really exist—or at least, not in a way that made people eager to pay $499 for a fix. Sure, we all complain about screen time, but when it came down to it, most people weren’t desperate to swap their phone for a brooch that talked back.

  • Tech Wearables are hard: Wearables are one of the toughest hardware categories. You’re asking people to change their daily habits and put a new piece of tech on their body—something that needs to work perfectly and look good doing it. Most wearables that succeed (like smartwatches or fitness trackers) solve very specific, daily-use problems. Humane’s Pin aimed to do everything, but ended up doing nothing well.

Why It Matters: 

  • First impressions are everything. If your version 1 flops, customers won’t wait around for version 2.

  • Real pain vs. imagined pain. People complain about phones, but not enough to switch to a wearable laser brooch.

  • Start small. Humane went big: massive funding, top-tier team, $499 product. Sometimes it’s better to start with something simpler and prove the concept first.

Trend

Le Chat and Grok

Last week, two big players made their next moves in the AI race: Mistral launched its chatbot app Le Chat, and Elon Musk’s xAI revealed its updated Grok-3 model.

Le Chat

Mistral, often viewed as Europe’s leading AI contender, rolled out its chatbot app Le Chat on iOS and Android. In just two weeks, it hit 1 million downloads—a strong signal that users are eager for alternatives to ChatGPT.

Mistral has built a reputation for developing cutting-edge language models like Mistral Large and its multimodal Pixtral Large. With Le Chat, the company is emphasizing both speed and flexibility. According to Mistral, Le Chat can generate up to 1,000 words per second, which is 13 times faster than GPT.

But speed isn’t the only selling point. Mistral is betting big on customization. Companies can deploy Le Chat on their own infrastructure, integrate their own data, and even tweak the model to better suit their industry or workflow. For businesses concerned about privacy or looking for AI systems that align closely with their operations, this kind of control is becoming increasingly attractive.

To strengthen trust, Mistral has also partnered with Agence France-Presse (AFP), a major news agency. This collaboration aims to ensure Le Chat’s answers are grounded in reliable sources—a notable move as AI models often struggle with accuracy.

Grok 3

Not to be outdone, Elon Musk’s xAI unveiled Grok-3, the latest version of its chatbot. During a live stream, xAI claimed Grok-3 is now ahead of OpenAI’s GPT-4o, Google’s Gemini, and other leading models in math, science, and coding benchmarks.

These performance claims haven’t been independently verified, but the exact numbers might not be the real story. What stands out is the speed at which xAI has improved.

Just months ago, Grok felt like an experimental side project built for Musk’s social platform X. Now, it’s positioning itself as a serious rival to the best AI models in the world.

Grok-3 also introduced a feature called 'Big Brain mode,' which gives the model more computational resources when tackling complex questions. This is somewhat comparable to chain-of-thought models like GPT-4o, which focus on breaking down problems step-by-step. However, Big Brain leans more on raw processing power to push for more accurate results when needed—a potential advantage for businesses that need both quick responses and deeper analysis depending on the task.

Alongside Grok-3, xAI introduced Deep Search, a tool that aims to improve AI-powered search. Unlike traditional search engines, DeepSearch shows users how it approaches a query—laying out its reasoning process before delivering an answer. It’s a step toward making AI search more transparent and reliable, something businesses relying on accurate data will value.

Why these announcements matter

  • The AI race is fragmenting: Mistral and xAI joining the likes of OpenAI, Google, Anthropic, and others shows that AI is no longer a two-player game.

  • Different models suit different needs: Le Chat emphasizes speed and enterprise customization, while Grok-3 focuses on reasoning and search transparency.

  • AI is turning into a business platform decision: Just like choosing the right software tools, picking an AI model could impact workflows, customer experience, and costs. The right model could streamline processes—the wrong one could slow you down.

  • Customization is a growing priority: Mistral’s enterprise setup hints at a future where businesses won’t just adopt AI—they’ll adapt it to their specific needs. For companies working with sensitive data or industry-specific tasks, this flexibility could be crucial.

Help Me Improve Failory

How Was Today's Newsletter?

If this issue was a startup, how would you rate it?

Login or Subscribe to participate in polls.

That's all of this edition.

Cheers,

Nico