It is not clear what the people of Humane.ai were thinking when they released their unbaked smart pin for reviews at the world's leading technology websites. The opinion of the technology journalists of The Verge, WIRED and also influential YouTubers was unanimous: this product is bad, and there is no doubt that their opinions will also carry weight with investors later on.
Initially, Humane.ai gained a lot of hype and hot air: it was a company founded by two Apple expatriates and raised 230 million dollars, which promised a revolutionary gadget based on artificial intelligence that would replace our smartphone: it is a smart pin worn on the shirt, based almost entirely on a voice interface, And should naturally perform various tasks that until now we have performed with the smartphone and all thanks to artificial intelligence. And to complete the futuristic look and feel - the pin does not have a display screen, but a laser projector that projects the necessary information directly into our palm. The future is already inside. seemingly.
The talked about and intriguing pin attracted a lot of attention in the demonstrations it did at the last Barcelona exhibition, including a well-known Israeli technology reporter who was enthusiastic about it. But then, as mentioned, units began to be released for review to the leading technology websites and journalists in the world. They were less enthusiastic.
Embarrassing mistakes
The Verge devoted no less than six videos to the pin, in which, to put it mildly, they tore it apart. First, many simple tasks that any average voice assistant can do, such as reminders or an alarm clock - are simply not supported. Even slightly more complex tasks that are supported today in almost every smartphone based on Android or iOS, such as navigation and object recognition, are also not supported. Website recognition doesn't work as well either: the smart pin correctly recognized a Mexican taco restaurant and knew how to say it had good reviews on Google, but it awkwardly misidentified Ryde as competing company Lyft - only because its sign is also pink... Similarly, the block in the name of Beyoncé in English, made the pin freak out: The Verge reporter asked her to play the singer's Texas Hold'em, not only was the pin unable to find the song and play it, in the voice response announcing the failure, the dash in her name became a jumble of letters: U+00E9.
Alongside this, the pioneering artificial intelligence projectile also presents hardware problems: it heats up, to a level that is unpleasant on the skin, being a device that is worn on the shirt, and also, according to the reports of the reporters - it reaches such levels of heating that it simply stops working. In addition, it is impossible to transfer the existing phone number to it, but you have to use a special designated phone number and explain to your acquaintances that the strange number from which they receive messages is you...
And finally, one of the features that should have been a selling feature of the pin, is the ability for simultaneous translation. In demos in Barcelona it worked pretty well, but The Verge let a Japanese- and Korean-speaking reporter screw up the pin. And she definitely came across. Not only did she fail to translate or respond properly, she simply repeated the sentences in what the reporter described as a "slightly racist accent." That may have been an exaggeration, but the fact is that Humane.ai simply failed miserably at the task it was supposed to fulfill. And when it comes to a gadget that costs $700, plus a $24 monthly service subscription, it's too expensive not to work.
The Verge devoted no fewer than six videos to the pin - here's one of them
The reporters at WIRED, one of the oldest and leading technology magazines in the world, agreed with their colleagues from The Verge. According to them, the promise of allowing us access to powerful artificial intelligence engines directly from the lapel of the garment doesn't really make skin and sinew - it pretty much remains at the level of the bones.
WIRED were unrepentant about the quality of the answers, which were simply wrong - "The next solar eclipse will occur on April 8, 2024," the pin answered the question. Only the question was asked on April 10th... Similarly, the pin was asked to identify what was in front of the wearer (an empty table) and answered that there were granola bars there.
Another gripe about the hardware (although they overall liked the photo quality the pin offers via the built-in camera) was its main output device - the laser projector. It's not bright enough to read clearly in daylight outside, which is a serious fault. They also believed that the interface, which is based on hand gestures and fingers in the air in front of the pin, is also complicated.
In short, the product is currently not quite ready for everyday use by the average person, and WIRED also points out that the pin is currently simply not capable of performing simple tasks that any smartphone can perform with its eyes (or camera) closed - such as ordering a taxi or putting reminders in a diary. The company promises more features in the future, but as stated there, we are reviewing what is there now, not promises in Alma - and it is an expensive product, which costs like a high-end smartphone to eat promises.
"Worst product I have reviewed"
However, the criticism that also caused a stir on the Internet, was from the successful YouTuber Marcus Brownlee, also known as MKBHD. The title of the video he put out on the pin was: "The worst product I've ever reviewed."
Although he did not come to different conclusions from those of his colleagues in the written press, his video, which received five million views, sparked a discussion, also about the duty of loyalty of journalists and content creators: to the commercial companies that create the products or to their viewers and readers?
Daniel Vassalo, a former engineer at AWS, who probably doesn't understand journalism too much, wrote in response: "I find it pointless, almost unethical, to say that when you have 18 million followers (to Brownlee - n.l.). It's hard for me to explain why, but with great circulation comes great responsibility (paraphrasing the famous quote from Spider-Man). The prospect of killing someone else's fresh project reeks of indifference," Vassallo wrote on Twitter and added at the end "don't do bad", taken from the doctors' oath. He just forgot that Brownlee is a journalist and vlogger, not a doctor. To this, by the way, Brownlee responded to him in one line: "We probably don't agree on what my job is."
MKBHD's review
And here I have to agree with Brownlee - a feeling has arisen in the world and Israeli media that journalists owe something or work for the technology companies they cover, and this is not the case. Vassalo is right that with great circulation, comes great responsibility, but not towards the companies or bodies covered - but a responsibility towards our readers and viewers. And our responsibility, first and foremost, is to tell the truth.
If a particular product is bad, and has serious flaws or glitches like the Humane.ai pin, it's our duty to say so. The claim that Brownlee's conduct borders on "unethical" is false and stems from a gross lack of understanding of the traditional role of the press, which, precisely in the current era of marketing content and content on behalf of, should adhere to its principles.
Not only does Brownlee have no ethics problem, the opposite is true. He adhered to ethical conduct by expressing his honest opinion: Humane.ai is simply an immature product. Or as WIRED's reporter summed it up - it may be the beginning of a conversation, but as the first commercial product based entirely on artificial intelligence, it is a failure.