Things aren’t going so well for AI hardware startups.
After several years of development, startup Humane launched a $700 wearable product that relies heavily on artificial intelligence in early April. Ai Pin’s original pitch was that it would eliminate the need to juggle different apps. Its operating system can “find the right AI at the right time,” and can even play music, translate languages, and tell you how much protein is in a palmful of almonds. . And since it doesn’t have a traditional display, the Ai Pin was supposed to be a little antidote to screen time ills. Smartphones were becoming popular.
The pin was panned. WIRED’s Julian Chokkattu gave Ai Pin his 4 out of 10 score. Popular YouTuber Marques Brownlee praised the device’s hardware design, but still called it “the worst product I’ve ever reviewed.” Since then, the company has promoted the message that this is a replacement for your cell phone. Bethany Bongiorno, co-founder and CEO of Humane, has apologized to dissatisfied customers and some of her fans with an apology, assurances that improvements will be made, and a laser projected into the palm of her hand. He has responded meticulously on Twitter, including a video demo of his UI for a gadget that replaces a smartphone screen. .
Humane seems to have lost its way with the launch of its product, but that’s not all. The cheaper Rabbit R1 was sold for $200 as a generative AI “pocket companion” and initially generated a lot of excitement, but has now been described as “underwhelming,” “half-baked,” “baked,” and “unreliable.” Labeled. His Chokkattu in WIRED gives him a 3 out of 10, but some people have questioned how the device handles logins for external apps like his Uber.
Such early hardware failures are not unheard of. Many startups have developed and shipped lackluster products as a result of over-promising in their marketing. Competing in hardware is especially difficult in an era of tech giants where the ecosystem rules everything. Developer Ben Sandovsky speculated that this was partly due to Humane’s co-founders sticking to the “Apple Way” and struggling in secret. Like any big tech company, they spent years honing their singular product, but instead of billions of dollars in cash, they were funded with $230 million in venture capital. , he wrote in a blog post.
However, both Humane and Rabbit appear to have made other errors in judgment. They relied on the excitement of ChatGPT-era AI to gain early customers and save themselves from the gadget graveyard. Instead, they jumped on his AI hype train and went straight to a non-functional brick wall. We found that generative AI does not make the hardware less difficult.
expensive flop
“To create really good new AI devices, you need to understand both the hardware and the software. The question with some of these startups is how much of the software layer is just a skin? ” said MG Siegler, partner at GV. , a venture capital company affiliated with Alphabet.
Siegler said incumbent companies in the technology industry have an even bigger advantage because they can build using their own infrastructure and can afford to take a loss on repeated new versions of their products. Masu. While startups are building crude AI products from scratch, Meta, Google, Microsoft, and Apple are leveraging existing teams and services to put and generate AI assistants in infinitely wearable sunglasses. You can mass-produce mobile phones with built-in AI search or create specified products. The laptop has a key for his AI, and the tablet has a “very powerful” AI chip.
“Big tech companies can give a hardware product five chances, but a startup may only get one chance,” says Greylock Investments, who spent several years growing the product at Snap. House Jacob Andreu says. “The chances of any of these small companies getting future funding rounds after releasing a big flop are never good.”