This is fucking dumb, and I am absolutely ok with this. The fact that we are able to do these 'dumb' and silly applications is a good sign. It shows that the field has advanced to the point where it is cheap enough to do these 'dumb' applications.
Not only that, but this 'dumb' application has immediate value(gets more people to use google's service and thus display ads to them). Because of the value these dumb applications have, they could help drive AI research and development of general purpose machine learning hardware.
Advanced? Since the earliest AI research, it's hard to think of a time when games,
puzzles, and toys were not front and center. Whether writing a chess AI would or a humorous chat bot, a focus on "silly" applications has always been the norm.
I'm not against such applications and share your opinion that they are great. But it's not new. The field has always been this way.
The feature is technically interesting, but I don't care about Allo. There are enough silly messaging apps out there, evolving in their own silos. I miss the days of the XMPP protocol.
I have always trouble remembering where a message came in. I have SMS / iMessage, Messenger, Whatsapp, 5 e-mail accounts and a Forum Inbox to worry about. And since you can't exactly flag messages as 'to do' I have to answer immediately or risk forgetting to reply.
I have Messages, which i think is Google's... and i only downloaded it because the apps from Samsung, Verizon, and Facebook went to war in January (on my phone at least) and somehow I ended up getting texts in either the samsung app or verizon but never both, but group conversations in facebook messenger, but only the verizon app would actually allow picture messages to be viewable if i sent them to my mother.
I also have Allo, because a (<-- not plural) coworker and I were curious about it. I think it solved that they didn't have facebook but did have apple, or maybe not i don't know. I also downloaded Duo and haven't opened it since, and I had hangouts, but i'm so very angry that Google screwed that up yet again. I have zero confidence in Google products fulfilling any of my needs completely ever again, and I really used to be the Google-champion of my friends... sigh.
A possible solution would be to run all third party apps in containers and use computer vision (deep learning) to collect messages and notifications from their screens into a single place, if they don't want to cooperate officially. I don't know the legal implications though.
Another app that should run into a container and be post-processed by computer vision is the web browser. We need to contain the browser itself (more privacy) and keep the pesky ads out of our eyes in a way that is impossible to hack by clever scripting or ad-block detection. On the next round, ad publishers would have to create adversarial ads that fool the classifier which would be a whole lot more difficult.
A consumer-focused messaging app. There's also Duo for video calls and "Messenger" for SMS (on Android). There's also the older thing called "Messenger" that's still running on all those devices that no one ever updated.
On the desktop/browser, I guess there isn't a consumer-focused messaging app anymore. Unless I'm missing something obvious, Google apparently ceded that ground to others.
I'm beginning to realize why Allo is as goofy as it is. It's not actually a messaging app. It's a machine learning playground app with a feature to send stuff to other people occasionally. Allo makes more sense as a concept if you look at it that way, though as a product, that's a bonkers idea that basically no one needs. Hence the non-existent user base.
App fatigue is real... when Captain Kirk flipped open his communicator he didn't select which app to use when requesting a transport. That Google has several apps which compete and conflict with each other is absurd. The fact they keep migrating functionality in and out of them is moreso. Just pick a winner and be done with it. There's not really a competitive advantage to splitting your ecosystem into 5 different communication apps, google, there's really not.
> when Captain Kirk flipped open his communicator he didn't select which app to use when requesting a transport
The original TOS communicator was just a dumb pipe into the ear of a switchboard operator (the naval rankings Uhura bosses around, presumably.)
In TNG and beyond, there doesn't seem to be "operators" per se, but the device still has a Conversational UX. Less like a phone, more like the Amazon Echo.
Presumably, in the future there will still be individually-packaged suites of software (and even such suites existing as rival products); but likely, in something like a space-navy, everyone has got the same software, and it's all been centrally requisitioned and integrated into monolithic monster workflows that show no hint of its origin. I imagine personal "communicators", despite the CUI, would resemble today's phones much more closely, as there'd have to be a way to install your own software packages and then manage the resultant overlapping suites of functionality, without requiring some full-time engineers to put it together into a streamlined system.
> That Google has several apps which compete and conflict with each other is absurd.
Google is a bunch of independent teams competing and conflicting with each-other (to launch products, mostly.)
The jockeying would be a lot clearer if each team had a distinct marque, beyond just "Google" or "Google Research" or "Android" or "Chromium Project."
That's not entirely correct. Allo isn't suitable for consumers, because you can't be logged into more that one device at once (last time I tried it you had to actually reactivate it every time you switched device). Additionally there is no web interface.
Hangouts does have an enterprise story, but I don't understand what Google's strategy is. I'm not sure they have one.
"Starting to roll out today" - does anyone versed in traditional PR have a say in this new trend in slow roll-outs?
I can see how it makes sense from an engineering point of view but I just installed Allo to try it out and it hasn't rolled out to me yet.
I'm rather unlikely to remember to try again in a few days/weeks etc.
This has happened multiple times in similar scenarios. I dug out my Gear VR to try the new Oculus browser. I didn't get it for 4 weeks (only a particular interest in WebVR kept me coming back to see if it was there).
Surely this is squandering good press and good will?
In most client apps that obey backend-account feature-flags, the app will introduce (advertise) the feature to you itself when your account gets the switch flipped.
AFAICT, they don't expect this announcement to reach or sway anyone who's going to use the feature; this post is more there to serve as genuine news—i.e. it's written to provoke water-cooler "what Google is up to" conversations among developers. (It's on the Google Research blog, after all; what consumer would even see it?)
Just did, it worked pretty well. It reminds me a lot of Nintendo's thing where they automatically make a Mii based off a picture of your face, except better. If you don't like the sticker pack that was outputted, you can fix your hair or whatever and try again. Took about 10-20 seconds for the first try, only 1-2 seconds for the second try, so there's probably some caching going on. My only complaint is that it's not super intuitive how to actually get to this feature. You have to be in a chat, go to stickers, go to the add stickers section, and then it'll be in that menu.
First I have to use a phone number to use an internet based chat system, now I should be excited about some comic figure? Allo - Are you out of your mind?
Not only that, but this 'dumb' application has immediate value(gets more people to use google's service and thus display ads to them). Because of the value these dumb applications have, they could help drive AI research and development of general purpose machine learning hardware.
Sometimes it's ok to do 'dumb' things.