Humane AI Pin review: not even close
theverge.com
theverge.com
There are broadly two use-cases of phones: content consumption, and utility. The former sucks time to increase "engagement" and has a bunch of social issues around it, the latter does not.
Both of these devices are trying to reduce the time we spend on our phones by addressing the utility aspect, but that's the wrong piece to focus on. The utility aspect is things like checking the time or calendar, booking an Uber, getting food delivered, etc. Many of these utilities however are essentially very optimised checkout funnels. Booking an Uber is really easy, because Uber are incentivised to make it easy. Same with ordering pizza. Taking a highly optimised flow and sticking voice on top isn't going to save time and is going to cause issues. These devices solve the wrong problem.
If a company really wants to improve the world by reducing phone usage (rather than just build their own platform to rent-seek on), maybe they should be looking at social networking, communication, federation, etc.
Being allowed to replace the home screen with some app would be game changing for making the device feel helpful rather than like the little universe of chaos that is is. Let me access other things from the app switcher.
Alas, we will never be allowed to do that.
There are also shortcuts you can automate and use your voice to open apps from the lock screen.
Yep you nailed it with that, however what do you think incentivizing success on "social networking, communication" looks like? Because it looks a lot like finding ways to increase engagement and scrolling.
I left out federation because normal people don't even know what it is and if they did they wouldn't care about it.
The problem with social networking is that the only viable business model so far has been ads, and without the high-intent you get from (e.g.) search engines, you have to instead optimised for engagement and keeping people around to make it up in volume of ad delivery. At least with services the revenue isn't ads based, or is less so. There are some things in the middle like Netflix/Spotify who want you to consume more, and have some ads, but I suspect most people don't want to reduce their music listening, and are happy enough with the amount of TV they watch (at least relative to social networking where I think it's very common for people to want to consume less).
Sure, the tech exists to make this thing possible, but it requires an Apple-sized company to execute on it, and even then it’s utility would be doubtful
There’s been a lot of pulling back on smart speakers because they’re not quite where they need to be - and this promised far far more than a smart speaker from a tiny new company.
The question is: did they believe they own hype or know it’s a sham?
By the looks of it, this device uses 4g, and doesn't support 5g, so over time we might expect service to get worse as telecoms build out their 5+g networks and leave the 4g ones behind.
as of the ultimate vision, it is not yet clear, and it would be dumb for anyone to get into this "AI phone replacement" business. There are far better use cases tho, which are clear to execute, like Schiffman.
Similarly a little box pinned to your chest projecting a display on your palm showing the weather makes for a good tech demo, but ultimately there is no real world use case for this device.
Set an alarm,
Create a reminder,
Or even attempt some of the more advanced features like querying local business.
The fact that it can't even set an alarm or create a reminder? Like what the actual?
Solo devs are already doing way more advanced stuff than this, it doesn't even take a small startup.
What takes an Apple sized company is creating something that can access many other platforms, buy for you, put together bulletproof many step processes and execute on them safely etc. But most of these failures are like basic basic stuff.
I do not understand how they can release a product like this.
The Humane Pin is aiming at this market, but with far less utility, ability, stability. They say they want to minimise the time we spend on our phones, but I don't regret the 7 seconds it takes to google an actor in the show I'm watching, or the 20 seconds it takes to book an Uber. I regret the 15 minutes I spend scrolling reddit on the toilet, or the hour I spend watching YouTube before bed. That's not the market they're taking aim at.
So this can be avoided by intelligently reducing the dimensionality of this incoming data (potentially with AI) and/or increasing input bandwidth
It's fantastic for expressing complex requests that would otherwise be a lot of effort/time consuming on a less-information-dense interface.
But voice has a higher "floor" for ease of use. It's easier to tap a button to confirm than it is to speak "Yes, ok" out loud. It's also more socially appropriate in more circumstances.
The other problem is that voice is very low density as an output modality compared to the status quo, which are high-resolution screens. The amount of time it would take to express even the most basic of information (imagine speaking the HN home page out loud vs. reading it!) is pretty extreme.
Where this forms a bad combination are tasks where it's not realistic for the user to utter the full complex request at once, where the user must consult intermediate outputs to determine the next action. In that case you're in a really vicious scenario: the density of voice input is not really necessary, while the low-density of voice output slows the task down dramatically.
For example, think of a use case where you're booking a flight: "I want to see flights from San Francisco to New York".
It's not really possible for the user to fully define a booking in their initial voice input - the user would reasonably want to review choices, pick seats, etc, necessitating that the task be multi-step. Now imagine if voice was the exclusive modality - the UX would be positively painful.
> "that's the future of most/all consumer interfaces with AI, no?"
And this is why I disagree with this statement. I think the idea that voice is the dominant user interface is far from obvious, especially when it comes to AI systems.
LLM hype men tend to hand-wave around the complexities of this: "the AI will automatically pick the best possible flight for you! Why would you even want to review your choices?!" - which conveniently dismisses every area of weakness for this UX with "the AI will do it for you, trust it"... and time will tell but I suspect that will not work out that way.
That could be fine, it's probably more comfortable than using a complicated website at least for some people, but again this is a feature you stick in a phone app, not a dedicated hardware device.
And that is startlingly worse as a UX than the status quo.
An agent can apply some contextual filtering to improve the initial choices offered to the user - but they can do that in a GUI as well, and I would strongly argue that's better served in a GUI than via voice.
"Given your check-in time and transit from the airport, I think the following 3 flights make sense... [painstakingly list all 3 flights verbally]."
"Oh ok uh can you say the second one again? Was that out of JFK or Newark?"
... etc. Whereas a GUI is easy to parse and presents choices side-by-side in a way that's easy to compare.
This is the dissembling I'm talking about when it comes to some LLM proponents - the idea that the user having to perceive, compare, and analyze information just goes away, poof because the agent will just... make it no longer necessary.
It's a fundamental misunderstanding of these domains and use cases.
I'll generalize my prediction a bit more: an intelligent agent applying its contextual knowledge to a GUI is likely going to be overwhelmingly better as a UX than an intelligent agent that largely interacts verbally.
That just screams a culmination of the last couple of years of "AI" and I am kinda loving that this happened.
Maybe, just maybe, Investors will finally realize that this isn't as magical of tech as the companies like to claim and can't be shoved into every single thing hoping for magical results.
Am I being too hopeful?
I hope not but I would guess yes. It's insane to me that anyone even bought this thing. Their marketing couldn't even show it in a positive light. It got things wrong, it seemed pretty useless, and the whole presentation that I watched felt like they were being held captive and forced to do things. The thing is connected to T-Mob which, in my mind, is one of the worse networks out there. AND it's $700.
But they sold some. Why would people buy it? Who looked at any of that and thought "YES! Let's throw $700 away for this poorly developed toy?!"
I unfortunately kinda agree, this alone likely won't do anything. But given the high likelihood of more than a few of these projects getting a ton of money coming out and burning in a similar way.
I think it will eventually happen after getting a few bad products.
Siri works great for a lot of the basic use cases that didn’t work in the video. Reminders, notes, alarms, messaging. Honestly I’m guessing it would work just as well for some of the knowledge questions that they asked.
Except if you do that you get an Apple Watch. It tells the time too. It also counts your steps, tracks your heart rate, functions as an exercise/fitness tracker, does a better job of showing your notifications, can stream music from services other than Tidal, can actually show Photos and Web results and emails (although the screen is obviously tiny).
It doesn’t have a camera though.
So that’s $370 less for the device and $15 less per month? For something that already works way better? And doesn’t seem to run the risk of burning you and constantly turning itself off due to overheating?
Hey look! That’s saving enough money that you could actually buy a second Apple Watch and still come out ahead.
Wow this is a dud. I wasn’t expecting it to set the world on fire. I wasn’t even expecting it to be good. I did not imagine it would be THIS bad. The two things I use Siri for more than anything else or reminders and texting. And this thing can’t do either one.
It just doesn't or people would be using it and people wouldn't be as excited about LLMs as they are. Comparing Alexa/Siri style, it works if you know the right trigger words and how it expects you to ask the question, also just does that one thing you can't follow on and expand on it is just a completely different world to what an LLM agent can do.
Apple Watch and Siri are dead products for good reasons.
I don't think we should discount this. There's something to be said for a product that works in a specific way, and if you use it in that way, it actually works. I don't need Siri to divine my hidden meaning regardless of how casually or ambiguously I phrase something... but I do want to know the phrase that works 100% of the time, and then I'll use that 100% of the time, and I'll be happy.
What I hate is how Siri fails at even that basic requirement a lot of the time. It's the worst for HomeKit scenes: "Set the scene 'In The Basement'" bounces between working and failing with an inscrutable error, release to release. Sometimes just saying "In the basement" works, other times it lists all the lights in the basement and asks me what it should do. I have a "Bed Time" scene that starts lullaby music in my kids' bedrooms, but sometimes Siri thinks "Bed Time" really means "Good Night", and turns off all of the lights in my house instead. Which is super great when my wife is in the kitchen trying to prepare things and suddenly the whole house goes dark.
For me, the biggest problem with Siri and other assistants is the opposite of what you describe: The product is sold as "natural language", which really means that it fuzzily matches what you're saying to its discrete list of features, and that fuzziness is only partially accurate, leading to broken use cases some non-zero percent of the time. LLM's are likely to be even worse here... maybe you've used a certain phrase every time you wanted it to do something, then some model update comes along and it decides to do something else with this phrase, and you'll have to constantly re-discover what works every time it updates.
Shouldn’t need to be, but is.
We already know they’re better, even if you screw up the question in some nuanced way which would trip Siri up completely they often get the gist of what you want and provide the expected solution.
Prediction engine vs trigger term syntax
That statement is several red flags.
I still remember wondering how Clinkle raised so much money. Look how far we've come since then!
The bigger issue with this thing though… why does it need to be its own device? It’s not going to replace your phone so why not just have this all happen on your phone?
But a 3rd party app will always be less integrated, have less permissions than functionality included by the manufacturer.
And for all this AI integration wide access is pretty much required as you'd want it to access your photos, notes, all kind of apps, etc.
This way manufacturers would have too much leverage over companies developing that kind of AI, as they could always develop better features than them with their own AI agent.
I think Apple Watch is a pretty good example of that already. Third party watches will never be as good as Apple Watch just because Apple won't let them.
You're at such a disadvantage on iOS and Android that it's a fools errand to try and build that app.
For AI to augment the lives of to the extend the iPhone did it's essential for it to be always on listening and able to act effortlessly.
Only Apple, Google and major android makers can deliver this experience.
There is however a window of opportunity for a team with the right talent to get there first if they're able to build their own device in time.
Apple are too privacy conscious to send all the data up to the server so we need to wait till they can build chips to do that locally or figure out a way to bend their own rules enough that makes it seem privacy focused, they also have a much weaker ML team so there is extra runway there while they choose who to acquire to fix that.
Google while extremely strong ML team its too academia brained to productize AI currently so need to wait for them to solve that, they also just suck at shipping products in general. They'll get there in the end but it's safe to say they'll only get there once someone else has shown how it should be done then they'll just clone it.
You have about 3 years before Apple solves this, so if you get yours to market and succeed in that time you capture a segment of the market before that happens.
Seems to echo the general reception to the first VR headsets.
Here's an even more effective solution: stop using your smartphone! After I stopped using my phone, I realized it had almost no value except some diversionary value that was actually just a distraction in disguise. I still have to use it at times such as 2FA, but other than that I rarely use it.
Phones can be used as a tracking device, and are much more likely to be used as such since everyone carries them around.
Buying a phone also supports the phone industry, that I do not want to support. It supports more mining and the disposable electronics industry. (Computers are more useful and if you install Linux on them, you can make them last for much longer than a phone.)