What reeled me in was the idea that the AI in Amazon's Halo could analyze my tone of voice and, theoretically, make that information useful to me...
My tech wish list this fall is brimming with stuff I may or may not end up getting. As I write, it’s the eve of the fall Apple iPhone event, where there’s always something new I want. I expect I’ll want to upgrade my phone, for instance.
Tomorrow is also Amazon Prime Day (busy week!), and that usually means a chunk of change on the Mastercard. My interest is piqued by the newest Amazon Echo with a smart hub built in. The new Chromecast with Google TV sounds promising, too, but I notice, as of now, Amazon isn’t selling the latest Chromecast device on Prime Day; it only showed me the third-gen model. It would appear Amazon wants me to buy a Fire TV streaming stick instead…. Imagine that.
Every now and then I’ll take a leap into the tech unknown — that murky area between gotta have it and what the heck. I joined an Amazon waiting list a few weeks ago for early access to the Amazon Halo, available “only by invitation,” with the allure of a 35% discount (to $65). I did the same for the Echo smart speaker years ago, and that worked out well.
The Halo is basically a fitness tracker. Amazon wants to convince me it’s more than that so that I commit to the service attached to it. Included in my by-invitation-only deal is six months free of the Halo subscription. For that I get “body composition, tone of voice analysis, sleep & activity tracking,” says Amazon. The sub auto-renews in April at $4 a month. We’ll see about that.
I received my Halo and opened it over the weekend. It’s a Velcro-strapped band — no watch face, no apps — that attaches to a sensor housing. It charges with a supplied accessory that fits over the housing with pins that align for charging. All communication is done by app. If you want to know the time, you need to wear a watch, too, or check your phone.
I don’t need a fitness tracker; my Apple Watch handles those duties while keeping me from having to look at the phone for the time. What reeled me in was the idea that artificial intelligence could analyze my tone of voice and, theoretically, make that information useful to me. If not useful, at least it might be a peek into the future of AI and what it has up its sleeve. I was also interested in the sleep analysis function, wondering if that could that somehow be beneficial during sleep-challenged nights.
I doubt that I’ll find that info captivating enough to pay $4 a month. Amazon lists Halo’s features as body, tone, activity score, activity intensity, sleep score, sleep stages, labs, steps, heart rate, live heart rate (I hate to think about the alternative), calories, activity sessions, time asleep, time awake and sleep temperature. When you use voice- and tone-tracking, seven-day battery life collapses to two, which makes sleep tracking tricky.
To learn more about the tone analyzer, I clicked on the details tab in the app, linking to a cartoon with a snazzy soundtrack instructing me there are different ways to react when aggravated by, say, watching the TV news. “When we better understand what makes us sound the way we do, we can make the choices that improve our relationships — laugh it off or make it a teaching moment,” said the robotherapist. I have a feeling the people who need to hear that message the most wouldn’t spend $4 a month to have their voice analyzed.
I measured my tone on Sunday and Monday. My most negative tone? At 9:30 p.m., just about the time the Dallas Cowboys went ahead of the New York Giants and my chances of winning the football pool dimmed. My highest energy moment (rewarded by a smiley face) came at 10:40 p.m. That might have been when I was taking the trash down and met my neighbor in the elevator — pretty much the extent of a party in socially-distanced times. I had just one phrase (really?) that sounded “happy, amused or friendly,” though Amazon didn’t tell me what that was. My most positive time was 7:21 p.m. Well duh… dinnertime!
Monday’s most negative moment occurred at 4:09 p.m. Now I know this isn’t true because I looked at my watch when I knocked the cheese grater off the kitchen rack and it grated my thumb on the way to the floor at 4:40 p.m.
The report showed I had two phrases in the early afternoon that sounded “confused, uncomfortable or skeptical.” Well, I’m a journalist — we’re supposed to sound that way. Overnight, I had four phrases that sounded “sad” and “restrained” at around 2:40 a.m., Halo reported, leaving me to worry whether I was awake or if I had talked in my sleep. That explains my rating of 65 (fair) for 5 hours, 21 minutes of sleep. It took me 18 minutes to fall asleep, which was interesting, and annoying.
I walked 9,106 steps Sunday, Halo said. Apple Watch said I walked about 7,200 steps. Obviously, Halo won the job as step tracker. I don’t recall giving my height to my Apple Watch, but Amazon had me do that during setup. Maybe that led to better gait gauging.
AI’s intimacy with my data is an issue for me. Giving my height and weight to Amazon — and yet more access to my voice beyond Alexa – veered outside of my comfort zone. I drew the line, fast, when Halo wanted me to strip to intimate apparel and shoot off a photo of myself for body fat calculation. I have plenty of body fat, and I know where it is. Thanks, but no thanks.
I’m happy to report that Monday’s shuteye time improved to 6 hours, 23 minutes, and it only took me 10 minutes to fall asleep (hmm, this could be addictive). I think I’ll stick with Halo for a bit. I’m curious to know how my tone registers when I’m doing an interview – do I sound as catatonic as I feel during allergy season?
If I decide to nix the subscription, I’ll lose the tone monitoring, but by then I think Halo will have done its job. I’ll still get readings for sleep time, heart rate and the oh-so-superior step-tracking. Hear that Apple?