MATH: Apple sold >250M iPhones 2008-2012. CIRP recently estimated the total number of Alexa-enabled devices at 31M. That’s what, .0018 apps/iphone vs. .0016 skills/Alexa device? This is a major problem? (4/14)
Good thread, thanks. Two thoughts: (1) VUIs are the command line interface problem all over again. We're back at the DOS prompt - you have to know what to say (command) and how to say it (syntax) before you even start to interact. Cognitive load is very high.
-
-
Thx, Dave! And yes, it’s back to recall > recognition. 1 in 4 voice app users said they couldn’t recall the commands. Most voice app creators haven’t been doing a good job of capturing & including all variations. It doesn’t need to be that way, but that’s where we are right now.
-
Another difference is that voice services should NOT be trying to include all variations of spoken commands. This is where AI is the critical differentiator - people want the same thing in different contexts, and different things in the same context. We CAN'T plan for everything.
-
Completely agree about AI being the critical differentiator. But until intent extraction gets better, VUI devs are stuck with commands. Unlike command line tho, they can easily add many variants for the same command. And I guess that's where I disagree: for now, IMO, they should.
-
Agreed. Offering a few alternates / variants is reasonable, but they should not be trying to be comprehensive. We need to wait for AI for that to be reasonable.
End of conversation
New conversation -
-
-
(2) Many skills or actions are not being developed by "individuals w/ dollar sign eyes", are they? The practical skills and active are coming from manufacturers and service providers (e.g., speaker and pizza makers), so they have a product or service behind them, not a fart app?
-
No, you’re right, and that’s my point: VUI app creators have little $ incentive, which is why we aren’t seeing the gold rush mentality leading to a market flooded w/ get-rich-quick app ideas. And why we aren’t seeing a flourishing of great VUI apps either.
-
I think there are fundamentally different task / behavior / goal sets between GUIs and VUIs. The overlap is where we are today ("what GUIs can effectively become VUIs?"), but new VUI roles are still emerging.
-
Some things can be more easily and effectively done by voice, but not all. GUIs are still better in many situations, so we it's not surprising to see the rush into services where VUIs are known to work well enough. Only a few will risk new VUI models, then we'll see the "me toos"
-
Yeah, the near term value prop for VUIs is more about context than ease of use (though they do well with basic queries & streamlining simple multi-step commands—e.g., play music, set timer, answer questions). They will supplement, not supplant GUIs.
-
As AI improves and their contextual data improves, they will become more like true assistants, and our need for visual feedback in many cases is likely to decrease. But we will always need GUIs.
-
I try to remind ppl not to fetishize tools/methods/emerging tech. As I have taken to saying, users don’t want to talk to their toaster, they want toast. Most ppl don't care about the UI, they care about their goals. Give them multiple ways to get it, but give them the damn toast.
-
GIVE ME MY TOAST! :) I say essentially the same thing - people don't want to use the tool (or think about how to use the tool), they have a goal and want the outcome.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.