Interesting article. Basically, Kik started off as just-another-mobile-messenger, and they had to compete with Facebook, Google, SMS, iMessage, WhatsApp, Blackberry, Skype, Snapchat, etc., so they figured bots would eventually set them apart.
Since all links inside their app open in their embedded browser, they can track click-throughs like Google, Facebook, and presumably the quirky 'bot interface' will be popular with people who are already used to chatting. No wonder everyone else is trying to maintain feature parity with them.
I'm sorry but Kik didn't predict the Rise of Chat Bots. Kik basically looked at WeChat and said we should do that. Same as Facebook and the others. WeChat showed that micro-apps inside a chat interface worked for their target market. The key is to understand the needs of WeChat's target market and figure out the overlap with the US market then figure out the unique needs of the US market.
The leap forward wasn't the tcl code quality, but an order of magnitude forward in what bots allowed to be possible in providing security to irc channels along with a host of other solutions simply was unparalleled.
The thing is, the Mobile Internet is rising, most people have phones, and old trends and concepts that are old to the Internet seem new and significant to the mobile platform. For example Flappy Bird was an ancient game on the web that existed and then it blew up, now you have many, many clones.
Edit:
Also consider how much information is on your phone, and how easy it is to access it than with a browser on your laptop. With the click of a button you give an app access to your contacts and other information. A chat bot you install wont really have this access, even if you gave it access, you wont normally keep your contacts on your system, not unless you use a Windows phone and it syncs to your Windows account (assuming this is a "thing" they do).
>For example Flappy Bird was an ancient game on the web that existed and then it blew up, now you have many, many clones.
I'll give you a better example than that in Angry Birds. Crush the Castle was a staple of mine for at least a year or two LOOONG before Angry Birds showed up; ultimately the exact same format of a game except your projectile was a trebuchet and not a slingshot. Angry Birds showed up on the smartphone and ate Crush The Castle's forever remaining lunches.
Hell, if you want to go back farther than that I think it was a game called 'Armor', fire at thing, destroy thing, upgrade and fire again for points. But yeah, that first sentence nails it completely.
The problem with Kik's version of bots is that they will face the same problems as apps do right now. Discoverability of bots, 100 bots that do the same small thing vying for user's time, bot overload, etc.
What Kik seems to have is essentially a web browser with a chat interface overlay where instead of typing in a URL you "mention" (summon?) a bot and the bot then gives you a "button" (link?) you can tap to open the web content.
One thing they do get right though is the idea that the bot should disappear once done and not bother you, however keeping marketers from not abusing that is going to require them to restrict that via code (e.g. session timeout)
Ideally I would want a single conversational chat bot that I can summon, ask to perform a task (regardless of what service is chosen to perform the task), select from the options provided, have the task performed. The bot can select an app to perform the task (deep link) or if no app can do it then some bot (if a bot exists to do that task) or a web link.
In some sense, twitter is IRC: you address @persons that meet on the same place you are, talking about some #topic. At one point I had the idea to actually write an IRC server where you would join a #channel that was actually a #topic, and the server would send you @people's tweets as messages. The usual guidelines for IM (putting @person's name before the message) would even be enough to reply to someone's tweet.
Couldn't agree more. Twitter is the IRC paradigm with a different presentation of messages by user in all their channels/#hashtags, vs in each channel.
> Pad (general-purpose computer message board), 1973, the first such, and a few months later, system-defined Notesfiles, precursors to Unix Newsgroups, Digital DECnotes and Lotus Notes.
Kik's a chat app and only a chat app (ie. they don't have a bunch of other properties that receive traffic like Google or Facebook) so this is what they've got, and what they have to work with. They can use them to keep people within their app and track where they go to afterwards.
They're gambling on the fact that bots will set their platform apart from all the other chat apps. Of course, now, everyone else also has bots. But the article is about how they did it first (out of the "western" chat apps).
I use a bot in Telegram called Nate. It's in a group chat and does all sorts of useful things, from Urban Dictionary lookups to wolfram alpha evaluations to spamming cat pictures.
I think there is a two part thing going on here, besides a lot of PR hype cycle echo chamber.
a) The anticipation that AI will be good enough to understand an instruction set through sentences. In some cases it already is.
b) Text isn't so much the interface but audio-voice interfaces are. The stuff from A) combined with voice recognition.
It is somewhat weird because Siri has been around for a while, Google voice recognition/search has been around for a while and been really good. Maybe the big aha moment was Amazon letting users actually input commands like ordering a product rather than getting dumped on search results for the web page to do it.
If the whole chat-voice-smart assistant thing is a winner takes all market, that winner probably replaces Google in mass market use cases for search. That is a really big deal and companies are salivating over the possibility.
Perhaps another factor is because other companies don't really have a knowledge graph like Google's, they are letting third party developers plug in.
So, the big tech companies have their products commercially available already and are being used by lots of people, the small tech companies have stuff they have been working on and are joining the hype cycle in hopes of being acquired.
So in summary the larger story is "chat bot" really = AI voice assistant which is plausibly the next Google eclipsing platform, and about as massive of a deal for the tech industry as the smart phone and self driving car.
So you think people will use voice to do general informational queries like what's a cure for asthma? Not just navigational queries like how to go to the theater?
This is the theory. And I don't think you're wrong at all in analyzing the motives/endgame here.
But there are reasons texting arose as a preferred alternative to voice. Audio interfaces will win in cars, possibly homes and other well-defined private spaces. But text is ubiquitous and visual interfaces are so much faster, as audio is limited by its linear dimension.
No doubt, voice control will be a huge market, but I think the idea is more alluring than the reality will prove.
36 comments
[ 3.1 ms ] story [ 88.7 ms ] threadSince all links inside their app open in their embedded browser, they can track click-throughs like Google, Facebook, and presumably the quirky 'bot interface' will be popular with people who are already used to chatting. No wonder everyone else is trying to maintain feature parity with them.
The leap forward wasn't the tcl code quality, but an order of magnitude forward in what bots allowed to be possible in providing security to irc channels along with a host of other solutions simply was unparalleled.
I mean, especially on IRC, with Eggdrop. My favourite was always: https://github.com/jamesoff/bmotion
Example: https://github.com/jamesoff/bmotion/wiki/Examples
Edit:
Also consider how much information is on your phone, and how easy it is to access it than with a browser on your laptop. With the click of a button you give an app access to your contacts and other information. A chat bot you install wont really have this access, even if you gave it access, you wont normally keep your contacts on your system, not unless you use a Windows phone and it syncs to your Windows account (assuming this is a "thing" they do).
I'll give you a better example than that in Angry Birds. Crush the Castle was a staple of mine for at least a year or two LOOONG before Angry Birds showed up; ultimately the exact same format of a game except your projectile was a trebuchet and not a slingshot. Angry Birds showed up on the smartphone and ate Crush The Castle's forever remaining lunches.
Hell, if you want to go back farther than that I think it was a game called 'Armor', fire at thing, destroy thing, upgrade and fire again for points. But yeah, that first sentence nails it completely.
Can't wait for someone to "innovate" mIRC's scriptable client into their product.
https://en.wikipedia.org/wiki/Eggdrop
Slack has effectively re-created irc and bots in a modern client.
It's nice mobile messaging platforms are joining the past and present with the future, let's just not call it innovation.
What Kik seems to have is essentially a web browser with a chat interface overlay where instead of typing in a URL you "mention" (summon?) a bot and the bot then gives you a "button" (link?) you can tap to open the web content.
One thing they do get right though is the idea that the bot should disappear once done and not bother you, however keeping marketers from not abusing that is going to require them to restrict that via code (e.g. session timeout)
Ideally I would want a single conversational chat bot that I can summon, ask to perform a task (regardless of what service is chosen to perform the task), select from the options provided, have the task performed. The bot can select an app to perform the task (deep link) or if no app can do it then some bot (if a bot exists to do that task) or a web link.
https://www.npmjs.com/package/@kikinteractive/kik
I mean, if it were actual "junk" (you get what I mean), I mighta be happy, but it's all "come to my private webcam site" kinda crap.
This is, maybe, why Whatsapp is so hugely successful: the cost of SMS/phone verification is too high for scammers.
Kik invented Chat bots, even tho IRC had them way before.
Your next startup invented xyz, even tho IRC had them way before.
> Pad (general-purpose computer message board), 1973, the first such, and a few months later, system-defined Notesfiles, precursors to Unix Newsgroups, Digital DECnotes and Lotus Notes.
> Talkomatic (text-based) (6-room, 5-persons-per-room real-time chat room), 1973, precursor to Instant Messaging Conferences.
Plus, emoticons in the 1970s! http://www.platohistory.org/blog/2012/09/plato-emoticons-rev...
To me, chat bots seem like a step backward, at least for the use cases I've seen them applied for.
They're gambling on the fact that bots will set their platform apart from all the other chat apps. Of course, now, everyone else also has bots. But the article is about how they did it first (out of the "western" chat apps).
a) The anticipation that AI will be good enough to understand an instruction set through sentences. In some cases it already is.
b) Text isn't so much the interface but audio-voice interfaces are. The stuff from A) combined with voice recognition.
It is somewhat weird because Siri has been around for a while, Google voice recognition/search has been around for a while and been really good. Maybe the big aha moment was Amazon letting users actually input commands like ordering a product rather than getting dumped on search results for the web page to do it.
If the whole chat-voice-smart assistant thing is a winner takes all market, that winner probably replaces Google in mass market use cases for search. That is a really big deal and companies are salivating over the possibility.
Perhaps another factor is because other companies don't really have a knowledge graph like Google's, they are letting third party developers plug in.
So, the big tech companies have their products commercially available already and are being used by lots of people, the small tech companies have stuff they have been working on and are joining the hype cycle in hopes of being acquired.
So in summary the larger story is "chat bot" really = AI voice assistant which is plausibly the next Google eclipsing platform, and about as massive of a deal for the tech industry as the smart phone and self driving car.
But there are reasons texting arose as a preferred alternative to voice. Audio interfaces will win in cars, possibly homes and other well-defined private spaces. But text is ubiquitous and visual interfaces are so much faster, as audio is limited by its linear dimension.
No doubt, voice control will be a huge market, but I think the idea is more alluring than the reality will prove.