Sunday, April 14, 2024

Humane’s AI Pin underwhelms at launch

Sunday, April 14, 2024

Helloooo, TechCrunch Weekend readers! I'll be taking over this newsletter from Connie. As ever, the goal is to highlight the in-depth, analytical, and opinionated stories that our writers are saving for the weekend. And hey — it's TechCrunch, it's the weekend, maybe we'll get weird once in a while.

Who am I? Well, I wrote for TechCrunch for many, many years, before foolishly departing to join a seed-stage venture capital firm. Now I'm happy to take the next step in my inevitable career trajectory: TC to VC to TC (again). Let's get into it! — Anthony Ha

Humane Ai Pin launches to tough reviews

Image Credits: Brian Heater

You can't criticize Humane for being unambitious: The startup's goal is to free users from their smartphones by building a wearable, AI-powered device without a screen.

It's a big challenge, and judging from the initial reviews of the Ai Pin (which launched this week), Humane hasn't cracked it yet. The Verge's David Pierce complained that the device "just doesn't work half the time," while Engadget's Cherlynn Low noted that given the $699 price tag, plus a $24 per month subscription fee, she "simply cannot recommend anyone spend this much money."

Some of the problems reported with the Ai Pin include slowness, overheating, and limited battery life — not to mention a general difficulty answering users' questions or fulfilling their requests.

TechCrunch reporter Brian Heater was more measured in his first impressions, acknowledging that the Pin is "very much a first-generation product in a number of ways" while also suggesting that it shows the "keen attention to detail" you'd expected from a company led by two Apple alums.

For its part, the Humane team says it's working to address reviewers' issues. Co-founder Bethany Bongiorno told Brian that the Ai Pin is the company's first product and that she and her co-founder have a "large roadmap" of what they want to build next.

Google bets on GenAI for healthcare and more

Image Credits: Getty Images / Nadezhda Fedrunova

Yes, we're talking a lot about AI this weekend, partly because (gestures broadly at the world), and partly because Google made a whole bunch of AI-related announcements at its Google Cloud Next event. In fact, Ron Miller notes that the company seems to be going all in on generative AI, with scarcely a mention of any core cloud tech at the event. Read More

Google Cloud is also collaborating with Highmark Health to use generative AI to personalize the patient intake experience. It's part of a growing trend of deploying generative AI in healthcare, but as the chief AI officer at the VA Sunshine Healthcare Network warned Kyle Wiggers, there are big risks in adopting the tech prematurely: "Relying solely on generative AI for healthcare could lead to misdiagnoses, inappropriate treatments or even life-threatening situations." Read More

How Neural Concept makes Formula 1 cars more aerodynamic

Image Credits: Williams Racing

Tim Stevens tells the story of Switzerland-based Neural Concept, which is using AI to create more aerodynamic parts for aerospace and automotive customers. Its core technology, called Neural Concept Shape (NCS), can suggest improvements or modifications to an existing 3D shape. Back in 2018, NCS was used to develop the world's most aerodynamic bicycle; now it's being used by four Formula 1 teams. Read More

A climate startup selling methane-eating microbes

Image Credits: Getty Images / Oscar Wong

For her latest Deal Dive, Rebecca Szkutak looks at Windfall Bio, which works with clients in methane-producing industries like agriculture, oil and gas, and landfills. Windfall Bio's microbes absorb methane emissions and turn them into fertilizer, which the client can then either use or sell. The Menlo Park–based startup recently raised a $28 million Series A led by Prelude Ventures. Read More

What else we're reading

Over at Platformer, Zoë Schiffer examines the debate over Jonathan Haidt's bestselling book "The Anxious Generation." Haidt argues that a "phone-based childhood … is making young people sick and blocking their progress to flourishing in adulthood." Other researchers have suggested the data isn't there to support Haidt's claims, while Schiffer tries to strike a balance between the established science (not definitive) and "the lived experiences of so many everyday smartphone users" (who experience negative feelings after spending too much time on their phones). Read More

Also, while Adobe's Firefly was supposed to be an ethical and "commercially safe" alternative to AI image generators like Midjourney, Bloomberg found that Firefly was actually trained on images from Midjourney. Oops. Read More