On April 14th, the tech YouTuber Marques Brownlee — MKBHD — released a video titled “The Worst Product I’ve Ever Reviewed … For Now”. It was among his most controversial videos, with some Silicon Valley types — namely SmallBets founder Daniel Vassallo — loudly complaining about his review. It wasn’t because they thought he was wrong, or were defending the product in question, the Humane AI pin, but because they said a notable YouTube reviewer, with close to 20M subscribers, has a huge amount of power, and a negative review can kill a company. 

As Mr. Vassallo put it on the platform formerly known as Twitter: “Hard to explain why, but with great reach comes great responsibility. Potentially killing someone else’s nascent project reeks of carelessness. First, do no harm.”

The core example of this power, which these critics cite, is the electric car company Fisker. Two months ago, MKBHD released a video titled “This Is the Worst Car I’ve Ever Reviewed”, and shortly after, the company’s stock tanked and they were delisted on the New York Stock Exchange. They started selling 2023 cars with  a massive, $24,000 discount, taking their Ocean Extreme from $61,499 down to $37,499, and bankruptcy seems looming.

But even for someone as influential as MKBHD, bad reviews do not kill companies; bad companies and bad products do.

Fisker released a flawed, incomplete car and because of this, received negative reviews; but their stock tanked independent of this, because they have a fundamentally bad business. At this point, it’s not that capital intensive to start an electric car company — Fisker doesn’t have a dealer network, and doesn’t run its own factories, instead outsourcing that to Magna Steyr’s Austrian plant.

That means that, if cars fail to sell or if there are more expensive issues than predicted, the company doesn’t have cash on hand, or assets against which to take loans. Fisker’s only collateral are the cars that aren’t selling; and even if all the software flaws were buffed out, it’s still a hard sell for consumers to drop over $30K on an electric car from a new company, with no track record, that is heading towards bankruptcy. I like the Fisker Ocean, but it’s not surprising that the business isn’t doing well.

Turning to Humane’s AI pin — that “worst product” Mr. Brownlee had “ever reviewed” —  you don’t quite have the fundamental business problem, but the product is even more difficult. On the business front, Humane raised over $241 million in venture funding, and barring outrageous spending on their behalf, they should have a lot of that left to cruise on – to improve the software, get a wider use rbase, and gradually win over curious customers in what will be an initially slow launch period.

But their product is where the flaws are. The Fisker has flawed software, but is fundamentally a quite good car, and cars have an obvious, usable purpose. 

By contrast, Humane’s AI pin is either a less convenient, less capable replacement to your phone, or an extremely expensive accessory that you won’t use. For those unaware, it’s a small square device that clips to your lapel or shirt and by tapping it, you can ask it questions, send messages, or ask it to take photos and videos of the world around you. Using a variety of connected AI models, it then tells you the answer aloud, or projects it onto your hand with a small, onboard laser projector; and the idea is that by wearing the pin, you can be more present in the moment. 

It’s a beautiful, very cool little piece of tech, and as someone eager to reduce their screen time, it’s right up my alley. This could be the future of mobile computing. But this isn’t there yet. 

Namely, responses to even basic questions can take over 10 seconds ; it can’t set a timer or alarm, or do many other basic features available on the first iPhone; it overheats, the camera is bad, you can only play music through Tidal, the battery life is extremely inconsistent and short lasting, the videos can only last 15 seconds, and though it can send messages, actually getting it to do so is a hassle. You have to tell it the message you want to send,  explicitly confirm which ‘Seth’ in your contacts you want it to text; and when it does, the message is from its own phone number, not your usual number.

This also means that, when you ask it to summarize your conversations, it can only tell you about conversations sent to its number, not to your phone. There’s nothing it can do that your phone cannot; and your phone will do it a lot faster. Also, you can only communicate with the pin by talking to it, so hopefully you won’t need to message anything private on it. Yes, you can technically type out a message using gestures on the projector, but that is so fiddly that it would be a pain to send ‘Yes,’ or ‘No,’ let alone any text longer than that. 

All of this for a $700 starting price, plus a $24 per month subscription.

In short, if the company fails, it’s not because a YouTuber was mean about it. It’s because nobody should buy their product.