The Ai Pin is worse than most imagined

The Ai Pin device that created so much buzz last year when its company, Humane, previewed it at a TED conference, has now reached the hands of tech reviewers and the their verdict is universally a thumbs down. As I wrote in my column at the time, the product was destined to be a failure, but it’s even worse than anyone imagined.

The device clips to your shirt or blouse, and you use it to make voice inquiries, phone calls, send messages, take pictures and video, and use AI to answer your questions and help with other chores. The well-funded company positioned itself as Apple-like, in part because its many ex-Apple employees and its attention to its industrial design.

When I first watched the TED talk, I thought the product was very strange, considering the company was trying to replace our phone, but offered few of our phone’s capabilities. It made little sense to me, was very expensive, and was trying to solve problems few of us have.

But because it was the first hardware product based on AI, it received a huge amount of attention and money. To date it’s raised more than $241 million from a range of companies and VCs, including Microsoft, LG Technology Ventures, Volvo Cars Tech Fund, Top Tier Capital, Hudson Bay Capital, Socium Ventures, Tiger Global, Qualcomm Ventures, Valia Ventures, Forerunner Ventures, Lachy Groom, and Sam Altman, founder of OpenAI. Was I missing something that these experienced investors were seeing?

A number of tech journalists have had the product and they are now writing their reviews. They are universally panning the product, mostly because it doesn’t work much of time. It frequently overheats, rapidly consumes batteries, and returns answers that were often wrong – a known defect of AI technology called hallucinations. Even it’s clever tiny laser projector that displays a green image on the palm of your hand is too dim to be used outdoors

Most of the reviewers described the same issues, but several tempered their findings with comments from the company, saying they are aware of the problems and software improvements are in the works. Let me tell you, these issues are well beyond being able to be fixed with a software update. Many of the problems are fundamental hardware deficiencies.

None of the reviewers recommended purchasing the product, certainly not at the $799 plus $24 per month for phone service. Marques Brownlee, one of the most influential and capable product reviewers, said it was the worst product he had ever reviewed.

What I can’t fathom is how supposedly smart people could ship a product that fails to work and does so little of what is being promised. It seems like corporate suicide, but having millions of dollars in funding sometimes distorts reality. The executives clearly know of these issues and shipping it seems like malpractice, perpetrating a fraud on consumers.

Before writing my earlier column I spoke to one of their investors, a former business associate, and he said the excitement for investors was because this was the first use of AI in a hardware product, and the potential it offered for many new capabilities. For example, they envision using the camera to look at a restaurant and describe its menu and ratings, or look at a plate of food and assess its nutritional value. The lure of AI and the rush not to be left out created the attraction for the investors.

While having a vision is important, it’s the easy part. The hard part is trying to turn these imagined benefits into reality, and the company has failed miserably. Most of us were skeptical of what the Ai Pin was supposed to do when it was first announced, and now we know it doesn’t even do those things well.

by Phil Baker