November 13, 2023 archive

AVFTCN 026 – Humane’s “AI Pin” gets closer to the Star Trek:TNG Communicator

5:09 am

From 1987-1994, Star Trek: The Next Generation (ST:TNG) brought an updated version of the classic show to our TV sets. Through my later years in university and first years out on my own, I tuned in each week to see the latest saga of Captain Picard, Commander Riker, Data, Worf, Deanna Troi, Geordi LaForge, the Crushers, and so many more.

Set 100 years after the original Star Trek series (aka “ST:TOS”), there were many advances in the fictional technology. One that captivated me (and probably many others) was the evolution of the “communicator” from a handheld device like a mobile phone… into a “badge” that was worn on your left breast.

You simply tapped on the badge and a whole realm of possibilities opened up. You could ask the computer for any information you needed. You could locate someone. You could call someone. It could serve as a universal translator between languages. It could do so many things!

It was so cool… and I always wanted one!

Over the years, several companies have made versions of a “communicator badge” that were Bluetooth devices that could connect to your phone. Essentially a wireless microphone/speaker that worked with your phone. They were cool… but you still needed your phone.

Enter Humane’s AI Pin

Until last week…

.. when a company called Humane launched their “AI Pin” device.

MANY articles have been written about it (and even more .. and still more).

At a high-level, the basics are that it is a $700 device that you wear on your chest or lapel and you speak to it … basically like that Star Trek:TNG communicator!

For a $24/month subscription, you have access to various services that it can perform, including responding to questions, playing music, doing translations, summarizing info, etc.

Additionally, it has a camera that can record images or video. An example they give is holding up a food and asking “can I eat this?”

It connects to all sorts of AI-related services, and uses large language models (LLMs) and natural language processing (NLP) to help with the voice interaction.

Plus, it has this cool laser display system that can project information onto your hand, and also allow to perform other gestures.

If you have 10 minutes, their launch video is interesting:

(And yes, loads of people have commented negatively about the video in articles … but regardless, it provides an overview.)

You just tap on it and start talking to it.

No separate phone required.

Now, it does require connectivity, and they mention that they have partnered with T-Mobile here in the USA. Each device has its own phone number and data connectivity. Effectively, it is a “phone” device as far as T-Mobile is concerned, even though it may never actually make a “phone” connection, doing everything over the data connection instead.

Always-on Connectivity

Which is, of course, given who I work for, one of the first things I thought about this device.

It will ALWAYS need Internet connectivity.

It can apparently do a few things in the device itself… but not much. Those earlier Bluetooth “Star Trek Communicator” devices needed the phone to be in range in order to do anything. The phone was the device that actually executed actions.

Similarly, if you use an Apple Watch or other similar smartwatch, you again need a phone in range for it to fully perform all its functions.

With this AI Pin, the devices need the network to be able to connect back to Humane’s servers and services. Otherwise you are just wearing a piece of metal and plastic on your shirt.

I do wonder how well this will work in many places.

I live in northwestern Vermont where our mobile connectivity is challenging outside of major roadways and our “urban” areas. (Our Vermont definition of “urban” may be different than yours! 🤣) You don’t have to drive far to find places with only very limited connectivity – or none at all.

But perhaps the main place these will be used will be urban areas, or within enterprises or other denser locations.

Or… if you look at the deal between T-Mobile and SpaceX… perhaps once they get the regulatory approval to launch new Starlink satellites that can provide “direct-to-mobile” connectivity, then these AI Pins will benefit from that other layer of connectivity. (Stay tuned… there are many steps before this can happen – and you can guarantee rivals will fight SpaceX every step of the way, so we may not see it for several years.)

The key point is that the AI Pin will always need connectivity – and presumably decent connectivity – to connect to its various AI services.

Subtracting functionality for a simpler interface

If you think about it – and this is a point Humane makes in its various presentations – we keep on subtracting functionality from our devices in a quest for the ultimate simple and minimal user interface.

Some of us can remember the ginormous mainframes and their terminals… which became mini-computers… which became desktop computers… which became laptop computers… which became smart phones… which became smartwatches and other wearable devices…

At each step, we’ve continued to remove user interface (UI) functionality. I’m writing this post on an iPad, which has less of a UI than my MacBook Pro laptop. I spend a lot of time interacting with people using my iPhone, which has a similar UI to the iPad, but just a smaller screen. I can also interact with people using my Apple Watch, which has an even more minimal UI.

The AI Pin takes this minimal interface so much farther. No screen at all (well… on the device… it can project via the laser thing). Really no UI on the device at all. Just tap it to do something.

Are we ready for voice-only?

Which raises the question – are we ready for a voice-only interface?

Maybe.

The past several years have been training us all… between Amazon’s Alexa, Apple’s Siri, Google’s Assistant, and all the others… we’re getting used to voice commands. How many millions of Amazon Echo devices are out there? (Side note… how many are actively being used?)

I think it’s also a generational thing. I’m a cranky old GenXer who deliberately DISABLES voice systems on all my devices. And I’ve only put one Amazon always-on-surveillance-device Echo in our house (where the primary use is to play music).

But I watch my 14yo daughter and how she interacts with systems. She uses Siri all the time on her iPhone. And using it, she can look up things far faster than I can type them into a search box. (Yes, we’ve of course raced each other.) She uses it on her laptop, and she uses voice all the time to interact with the Echo in so many ways.

My teenage daughter’s generation thinks nothing whatsoever of talking to devices.

The technology and systems have come so far since the ten years I spent in the voice-over-IP world back in the 2000s and early 2010s.

And with all the developments in AI systems, the voice interaction systems are so much better. It’s truly remarkable.

But… AI.. and accuracy?

But wait… how accurate can AI systems be? What about all the times when LLMs completely make up information?

Yes, this is a major issue.

In a completely shocking lack of fact checking, the Humane folks produced and published their launch video with a blatant error from their LLM. They asked the AI Pin when is the next solar eclipse and where is the best place to view it.

The system answered correctly that the next total solar eclipse is April 8, 2024. But it then answered falsely that the best place to see it would be Exmouth, Australia, and East-Timor.

Screenshot from the Humane AI Pin launch video showing the two founders wearing black standing in front of a table with their devices on it. On the right side this the text in green “The next total solar eclipse will occur on April 8th 2024. Best places to see it are Exmouth, Australia, and East-Timor.”

🤯

How do I know this is not accurate?

Because the part of Vermont where I live is directly in the path of the April eclipse! (In fact, some organizations are seizing on the opportunity to rent out homes to people who will visit us to watch the eclipse. 🙂)

The SF Gate site figured out that these names were the best places to see the 2023 eclipse. 🤦‍♂️ (They also pointed out that later in the demo when he uses the camera to figure out the amount of protein in some nuts, the amount is significantly off!)

🔥🔥 MAJOR FAIL 🔥🔥

And it blows my mind that someone would not have fact-checked something as major as this launch video. 🤯

Be that as it may, there is a huge amount of work happening to make these systems more accurate. Over time I think they will improve – it’s just a question of how much time that will take.

(And a point that right now would NOT be a good time for anyone to wear one of these if they need live-saving data!)

Privacy?

My other major concern is of course privacy. All this data is being transmitted from this AI Pin device up to various cloud services provided by different vendors.

Om Malik had a lengthy interview with Humane co-founder Imran Chaudhri where he specifically asked about privacy concerns. Chaudhri answered about how privacy and data protection were critical elements of their approach. And that this was in fact why they are only having limited partners initially – because they have very specific privacy and data protection requirements.

He spoke about people being in control of their own data, and about partitioning data so that even the company doesn’t have access.

I have no doubt that this is part of their core values… today.

But what if their company fails … or is bought by another bigger company for whom privacy is not a core value?

I know people with a Fitbit device who are greatly concerned about what Google is doing now that they own Fitbit. (And Google is in fact phasing out Fitbit in some parts of the world and pushing people to move to their Pixel devices.)

I do worry about how private any device like this can be in the long term.

Will I get one?

Not yet. But… I won’t rule it out.

First, the $700 is a bit steep for me to just have a device to play with – and I don’t really feel like adding another $25/month (or whatever it is with taxes).

Second, I don’t know that I could trust the info coming out of it yet. The blatant errors in their demo show that it’s not ready for major usage yet.

Third, privacy. Privacy, privacy, privacy. There’s a reason I disable voice systems on my devices. 😀

Fourth, there just aren’t that many services yet. That will come, but right now there’s not enough for my frugal Vermonter personality to justify the cost.

BUT… if someone gave me one as a Christmas gift, I’d definitely play with it to see how it works! 🤣

I can see the promise and the potential.

I mean… I watched 7 years of ST:TNG… and then ST:Deep Space Nine… and then all the other Star Trek series that had this kind of device.

I have wanted something like this for the last 35 years! 😀

We’ll see.

Will this AI Pin device actually prove to be useful? Will it be the launch of a new way for users to interact?

Or will it fail and just be yet-another-exciting-device-that-winds-up-dead-in-a-drawer?

I don’t know. I do know that it’s exciting to see something like this with a new form factor and services.

I look forward to seeing what Humane does with it… and if there are other companies and services that join in with this type of device.

Meanwhile… this went on far longer than I expected… and I need to go drop off my daughter at school and start my work day!


Thanks for reading to the end. I welcome any comments and feedback you may have. What do YOU think about this kind of device? Would you get one?

Please drop me a note in email – if you are a subscriber, you should just be able to reply back. And if you aren’t a subscriber, just hit this button 👇 and you’ll get future messages.

This IS also a WordPress hosted blog, so you can visit the main site and add a comment to this post, like we used to do back in glory days of blogging.

Or if you don’t want to do email, send me a message on one of the various social media services where I’ve posted this. (My preference continues to be Mastodon, but I do go on others from time to time.)

Until the next time,
Dan


Connect

The best place to connect with me these days is:

You can also find all the content I’m creating at:

If you use Mastodon or another Fediverse system, you should be able to follow this newsletter by searching for “@crowsnest.danyork.com@crowsnest.danyork.com“

You can also connect with me at these services, although I do not interact there quite as much (listed in decreasing order of usage):


Disclaimer

Disclaimer: This newsletter is a personal project I’ve been doing since 2007 or 2008, several years before I joined the Internet Society in 2011. While I may at times mention information or activities from the Internet Society, all viewpoints are my personal opinion and do not represent any formal positions or views of the Internet Society. This is just me, saying some of the things on my mind.