Chatbots were never the “next big thing” by any measurable metric other than artificial hype created by the companies trying to create a new platform for extracting revenue and data from users.<p>For this argument to make any sense, there would need to be data that showed growing usage of chatbots, followed by a plateau and then a drop off. I’m not sure what that data looks like, but I’m pretty sure there was never a high growth phase where users were actually interacting with the chatbots. It was all hype created by Facebook and a few idiotic VCs who wasted money on what they thought would be the “next App Store.”<p>I, and many others, were saying around the time of this hype that chatbots would never become a category defining product, simply due to inherent usability flaws in their design that have been discussed ad nauseum.<p>The only people surprised that chatbots failed to become the “next big thing” are the people who mistakenly thought they ever would be. This assumption was never grounded in any real data of user desires or real problems. Chatbots were then, and are now, a solution looking for a problem. I’m not surprised at all.
It’s just like why all voice interfaces are shitty: no one has any idea what the thing can or cannot do. They have a hidden user experience but the interface makes it feel like you’re talking to a human, but it’s so far from being a human.<p>These interfaces are almost like a dark pattern because of how bad they are.
I always found the chatbot idea odd, it felt like a step backwards in terms of interaction. We started out with very basic input methods to computers, like punch cards. Then we moved onto a command line interface where you could type in words. Then we eventually got GUIs, graphics, websites, all sorts of complex and nuanced ways to interact with a computer.<p>To go back to interfacing with a program using written language seemed like an odd step. It's never been the most efficient way of doing something, and it requires very advanced technology to accurately understand what people are trying to say, in whatever slang, shorthand, or bad spelling/grammar they use.<p>Besides, it's not really dead, the tech just moved to "voice assistants" rather than "chatbots" - really just spoken word rather than written word. And I'm not convinced that's the "revolution" most people are expecting either. I'll stick to clicking buttons and typing things into my terminal.
Whenever a company's phone is answered by an AI, it never does what I ask it to do.<p>At that point I just push 0 and repeat "speak to a human."<p>Why, oh why, would any rational person think that this kind of technology was about to suddenly take over everything?<p>This is a typical example of groupthink delusions.
1. "Chatbots" is horrible branding. Especially with the hype coming near the unfortunate timing of Russian bots interfering in the US election.<p>2. "Chatbot" makes me think the user is leading the conversation. The bot is interpreting what the human says, and then the bot tries to respond appropriately. This leads to customer service or operations (like calling a car) use cases. These specific use cases are hard to pilot, hard to get right and really tricky to show value to the user or the company. I've never seen a compelling problem/solution/use case for customer service chatbots. The math doesn't work.<p>3. Messaging is a real channel. It has advantages over phone calls, email, direct mail and even face to face conversations. Chatbots just aren't the right concept for the messaging channel. Dentist reminders work great for instance.<p>4. If you like thinking about this stuff, I do a podcast all about messaging. It's called The Chat Bubble, and this episode is a good place to start if you want to go deep on FB Messenger. <a href="http://thechatbubble.com/2017/11/edit-content-the-playbook-overview-the-4-top-marketing-use-cases-for-facebook-messenger/" rel="nofollow">http://thechatbubble.com/2017/11/edit-content-the-playbook-o...</a>
I think it's equally wrong to just write off chatbots. They are a User Interface tool that works in some areas and doesn't in others (and recently has been dramatically misapplied).<p>Chatbots work well as an input when your hands are otherwise occupied:<p>- driving/directions (Google Maps telling you where to go)<p>- cooking (reciting a recipe)<p>They work well when you're requesting a specific thing:<p>- "Play Everlasting Light by the Black Keys"<p>- "Add Eggs to my shopping list"<p>- Responding to Answers in Jeopardy: "What is Syracuse?"<p>They can work as an alternative CTA in certain narrow areas where they function like a traditional "wizard". I've seen some ecommerce stats for things like "Are you shopping for a Fathers Day Gift? Does your dad like sports? Does your dad like gadgets? Want to see some popular gift ideas?"<p>Where they don't work:<p>- complex NLP dependencies<p>- data entry<p>- when there is an expectation set that you're talking to a human.
Users never wanted chatbots, really. But platforms wanted them because it meant yet another way to ensure content is locked away inside their platform. So, platforms shout to the high heavens, and everyone rushes to adopt they realise there is no audience.
> An oversized assumption has been that apps are ‘over’, and would be replaced by bots<p>That was never going to be the case for sufficiently complicated tasks with enough decisions to make. Same reason Google Home/Echo are awkward interfaces for almost anything other than getting music to start playing.<p>GUIs on phones/desktop are tremendously productive because they're information rich and trivial to use. I can search, choose, customize, and purchase an item on Amazon in like 15-30 seconds flat. A bot isn't going to beat that for the general case.
For me, Chatbots felt like almost a violation of my privacy. Many marketing industry related sites would get your permission to send you an eBook, but would slowly after a couple of days drop you a random message about an upsell to their event/conference/webinar/coaching (some random high ticket value item) and it's terrible than email spam because email is not intrusive, but chatbots are VERY intrusive. Suddenly, my phone would buzz, my chrome would send me a notification and my watch would vibrate all at the same time only to find out that this shitty marketing firm had sent me something I don't care about. And the worst ones are the ones that keep sending you stuff with no way to opt out unless you resort to blocking them. These should be made illegal.<p>Another case study - I worked with a large pizza chain to implement chatbots onto their site. In the end, I (although I knew I'd lose money) told them that their basic add to cart, checkout experience on their e-commerce site was far better than a chatbot asking you 101 questions to make a f<i></i>g pizza.
I didn't advice them based on just intuition, we actually setup Google analytics and saw that the conversion rates were far better on the E-Commerce site and that it was impossible for the IT team to keep up with the marketing team that kept running promotions which led to tons of edge cases for their ordering flow.<p>So, another example of great technology, great potential, but destroyed by greedy marketers, poor implementations and businesses that simply want to ruin an otherwise perfectly customer work flow.
Chatbots have no ability to interpret unexpected responses, and are very infrequently trained correctly. My electric company has a good chatbot, because when I said:<p>"The electric pole outside my apartment looks like it's going to fall over"<p>It ended up using that sentence to determine that I should go to emergency on-call, which is all it needed to do; further parsing was irrelevant. But then I called Fedex this morning and told their chatbot:<p>"My package has a delivery exception."<p>And it simply couldn't wrap its head around the problem at all — apparently Fedex doesn't train it to recognize terms displayed to shippers! — and so I had to battle through the usual "operator, domestic, 1234" route.<p>If you have a chatbot, be absolutely sure to train it on every word your company uses for official purposes. The electric company clearly recognizes "pole" + "falling" as an emergency, but Fedex doesn't recognize "exception" because their mindset is "which division of Fedex Multinational Conglomerocorp do you need", even though they directly asked "why are you calling".
We've had a better alternative to "chatbots" for a long time, it's called a command line.<p>All of this obsession over "natural language processing" is annoying to me.<p>I have a Google home, and it can't do even the most basic things and struggles to understand whenever I tell it to do something.<p>It would be nice if I could just talk to my Google home as if it was a command line, with very explicit and specific commands that do powerful things. "Google take a note. Do the laundry." "Google play spotify on the living room speaker" etc. Instead the interface is an unhelpful mess that doesn't know how to help me get it to do what I want.<p>I feel that it is the same with chatbots. Stop trying to process natural language, and give me a command prompt with good documentation and user on-boarding.
I think all people with some technical insight could've told you that 3 years ago. But still I'm happy the world finally realizes it. Now it only(!) takes another 5-10 years until my employer figures it out. Some of these bigger companies just start to bet on chat bots, but of course will fail at the same hurdles as everybody else.
'Next big things' are never the next big thing.<p>What happened with Chatbots was the result of some very very poor 'ui translation' from Asian apps. The product designers and managers of the western world looked at 'mega chat apps' like WeChat and Line that have all these built in apps and interfaces and tried to replicate that. Instead somehow they thought they were textual chat interfaces (buy plane tickets in natural language) when in reality they all had their own GUI with buttons and things to press.<p>Then all the businesses and agencies and consultants that thought they could cash in on this opportunity saw it and started pushing it even harder.<p>There was never any real user problem that this was solving - just businesses trying to cash in.
Disclaimer, I work for a chatbot startup that is doing really well in the Nordics. In two years we have grown from 3 employees to almost 100. Chatbots are a real big thing, our business is stronger than ever, end-users and our customers are engaged in this! We solve real business problems, we integrate the chatbot with existing business systems. With this is mind, chatbots are not perfect, but most end-users are forgiving and accept this.<p>Chatbots are not a fad, its a real thing, and you will only see more of them in the coming years.
Actually, they may not be the "next big thing", but they're still a pretty huge deal. Especially for IT and ecommerce support. The support sector is estimated to a value of a few trillions of USD / year in the coming years, because of the progressive shift of large cohorts of customers to the cloud.<p>The point is not to have chat-bots handle all cases, but if they can handle 80% of cases, which are generally trivial, then it'll good enough.<p>Of course I hate talking to a bot just like anyone else, and being cornered into a dead-end discussion with one, or to be in an infinite loop, or in a situation where the bot has no answer and does not offer another path to resolution. But, from a business perspective, it's a rather sound approach to have people go through a bot first rather than have every single complaint clog an inbox. It doesn't scale so well.
I seem to remember these were the objections that were made at the time.<p>The basic tech dates back to SHRDLU and ELIZA. In some ways this is similar to the partially-self-driving-car problem: if it's a mechanical interface where the human is in control it's fine, and if it's a hypothetical human-equivalent AI it's fine, but in the middle people forget to adapt to the limitations of the system. And the limitations are very severe as soon as you go outside the lines.
There are/were two main problems with chatbots.<p>1) There’s a huge consolidation around Siri, Alexa, and Google. No one wants to download an app, and you’re never going to win against wake words.<p>2) Doing the NLU and intent modeling is really hard to get right. All too often, someone comes along, thinks all they have to do skis throw together some regexps or toss the neural net du jour over some text and it will all just work. It won’t. It doesn’t. It’s just really really embarrassing.<p>Intelligent back off is a thing. Understanding when you don’t understand something is a thing. Conversation management is a thing. UX matters.<p>Also, trying to fake chatbots with people, and then saying “We’ll just train up a model and get rid of those folks” was never going to work, because you’d always have to have people, because you set your bar at hard AI then. Sure, that’s a decent strategy of you already have a call center that you’re trying to cut costs on, but for a startup, that’s just hubris.
I'm the creator of a Slack bot (Pull Reminders). I ended up building a web UI to configure the bot rather than using slash commands. The main reason for this was that Slack's platform is still evolving and I thought it'd be difficult to build an intuitive experience/interface within Slack. I've seen other Slack apps do a pretty good job at this though. Check out Eventbot or GitHub's official Slack app.<p>It's been fun figuring out how to tackle UX problems like onboarding and engagement with Slack as the primary channel (versus email). I think chat is replacing email for many teams and so there's a big opportunity to rethink traditional workflows with a "chat first" user experience.
Recently built a chat/comms. system for my last company. I think key to this user experience is automatically routing to a human when AI systems cannot determine intent. Most cloud products in this space now provide some way to determine a confidence measure. IE the default should be human interaction, unless an intent is certain and a simple remedial action can be taken. This kind of workflow does have the ability to increase efficiency, and improve customer experience if done correctly, but it can easily go wrong.<p>For example, with my bank, automated telephone banking is often much more efficient and quicker, than talking to a human CSR/Banker, for moving money around in my accounts.
> On the input side, it’s easier and faster to click than it is to type.<p>I disagree with this. It may be easier, but it is in no way faster. For those who don't use computers very often I can see it being slower. But for people who use computers on a daily basis, for mice to still be a commonly used input device is disappointing. With constant context switches, moving an object across the desk with you wrist, while keeping your fingers in the same place to click, and carefully putting the cursor just in the right place isn't nearly as fast as the same interface using typing. Or comfortable for that matter.<p>I'm not saying I disagree with the article as a whole.<p><i>EDIT: typo</i>
Hype is nothing but unrealistic optimism and thats what the current world of tech is filled with. People see staged demo and the media start throwing wild speculations. I never ever get excited by any demo and thats what I suggest to people but I guess the whole stupid idea of "be optimistic" is so rampant in current culture that people just blindly start following anything.
I think we set the goals way to high. It seems like everyone expected "chatbots" to be in their final form right away. Currently, the best "chatbots" aren't really chatbots, but act of as sort of human backed, virtual assistant in a narrow but widening domain. For example, real estate. Most real estate companies have an insides sales agents that does lead qualification, sets appointments, etc. There's a company in the space doing this now and judging from their website, they are doing alright.<p><a href="https://structurely.com" rel="nofollow">https://structurely.com</a><p>I don't see one mention of "chatbot" on their page, it's always virtual assistant. Even just checking out their marketing material it looks pretty compelling. <a href="https://research.structurely.com/customers" rel="nofollow">https://research.structurely.com/customers</a>
To my mind chatbots are the great way to improve your marketing strategy. For example, our company provides different AI data analysis API’s for businesses, such as multilingual summarization, news aggregation, sentiment analysis, data extraction and etc.<p>We are using chatbot <a href="https://www.summarizebot.com" rel="nofollow">https://www.summarizebot.com</a> as a part of our marketing strategy. And this works. Potential clients in Slack and Facebook are using our chatbot and after getting interested in our SaaS solutions. The conversion rate in Slack is quit good since a lot of businesses are using it.
The strategy is being successful since out potential leads are able to see and use our main technologies in the bot. Chatbot for us is like smart demo of our existing technologies and features.
Having a chat with someone can be enlightening because the conversation can be engaging, can grow organically and the ensuing organic path can lead to insight.<p>Chatbots, to me, just don't feel very organic - it's very much a request/response paradigm, often with a limited set of useful responses (perhaps I have just encountered the basic ones). If that's all we're getting, I'd much rather just hit buttons on a screen, than type an entire question in the hope that the bot understands it (when it doesn't, it's just an annoying guessing game).<p>Plus, as other comments have mentioned, it's very hard to see what the bot is actually capable of. With a good screen based UI, the capabilities are much more obvious. If it's not fit for your purpose, you can just quickly move on.
I think for the most part, chatbots are just stupid. They're like the untrained outsourced call center employee, except that the untrained call center employee at least has 20 years of experience interacting with humans in a non laboratory environment.<p>It's in the uncanny valley where it seems almost human like but a major turn off because it doesn't act like one.<p>It's probably just better to have the bots act like robots instead of "chatting".
I'm one that's not easily swayed by hype usually.<p>I haven't jumped on the crypto bandwagon even, but I was working on a project midway through last year. It was a concept that would help people over come anxiety and depression so they could self-actualise and live a better life.<p>The thing with many people -- from out experience -- was that they don't necessarily always want to speak to a professional but they do want to share what they're feeling with someone they can trust. That's often a friend, but unfortunately friends aren't always in the best place to advise and sometimes inadvertently can make the situation worse. So we decided to build a bot that was anonymous enough to create enough abstraction but friendly enough to help. The content was driven by experts in the field.<p>Anyway. The whole concept just kept failing. We could barely get people past the first stage of questions. We tried so many different platforms, we did so many user studies and we kept trying to iterate but it just kept failing and all this time I did wonder in the back of my head if an app would have been a better idea. Now I know it would have probably at least been a better idea than a bot.
IMHO the thing to get excited about is the potential for a somewhat ubiquitous UI to get my stuff done; not the NPL. Hopefully NPL will continue to improve and be leveraged more in the UI. But right now, the prospect of ditching the seemingly infinite number of GUIs I have to deal with between the web and mobile apps for tasks/bots in the messaging platform of my choice is what's peeked my interest.
Chatbots solve the same problem as a search interface, but in a way that is lower bandwidth, less intuitive, and mildly personally insulting (the attitude from the company using the chatbot seems to be 'our users totally will not mind wasting time trying to interact with a really shitty set of canned responses'). The bigger question for me is 'why did anyone think this was a good idea?'
Chatbots are an infantilization. Just give me a list of commands I can type straight in and quit with the fake human bit.<p>Chatbots are basically command lines with some guidance for non expert users. Convenient when you don’t want to build a bunch of custom GUI for new features but often gives a worse user experience when the commands are dreadfully buried in long winded conversations.
Chatbots are the next big thing. The reason they didn't take over is because NLP and AI sucks (and "sucks" is putting it very lightly). Current AI is stuck in a rut, and they don't even realize it. True innovation in AI will come, and surprise everyone, and like all innovations before, it will come from cutting edge scientific research.
I'm unable to recall a single instance in my life where an automated system was more useful than talking to a human or having great documentation.<p>From phone directory systems to automated help-desk responses, I've spent more time in recent years trying to get around an automated hurdle to talk to a human than waiting on a forum response in 2004 or sitting on hold back in the days of 1-800 help numbers.<p>Chatbots were a predictable failure.<p>Maybe I'm just older now. Settled in my ways so to speak. But I've dedicated my professional life my tradecraft of media production and publishing. At some point the technology tools stopped focusing on improving life or solving the problem.<p>- The first VR movement was WAY too premature.
- 3D TV's were DOA.
- 4k+ / High resolution is a drastically misused/misunderstood medium and technology.
- The current VR movement just simply doesn't have a platform that can support the level of processing and bandwidth needed to be worthwhile.
- Apple has lost it's way.
- I don't even know what Google actually does now, but I cared more when the search engine was amazing.
- There's too much media, not enough curation.
- Blogs and personal websites were the golden age of great information
- RIP copy editors
- Social Media is a game of smoke and mirrors
- WHO IS CLICKING ON ALL THESE ADS?!? Nobody I know, yet the big players are making billions? Are the ads worthless? Are we going to wake up one day to see the internet crashing and burning the same way cable television is? Don't get me started on newspapers.
- AR is back? Why?!?
- I'd love for RSS to make a comeback, if there was anything left to subscribe to.
- AI seems to be the future but I've yet to be impressed. The things that are promising are far out of reach from everyday users that it just seems far far away at this point.
- Fortnight seems to have hit the nail on the head.
With the chatbot hype, there was a healthy number of skeptics throughout: people who wondered about the technical realities of delivering satisfying UX, ones who worried about unanswered questions about business models, and ones who figured out that this future would be most beneficial to gatekeeper-platforms who'd then act as discovery facilitators for users to select from among competing bots, rather than for botmakers themselves.<p>It's no accident that Amazon and Google are currently leading in the consumer voice assistant market: they had large, pre-existing base of users and enough intrinsic first-party functionality to bootstrap their assistants into rudimentary usefulness, but then they built platforms where third-parties could compete for users the same way it happened for apps. In the app boom, aside from a few runaway hits, the only ones who reliably got rich were the platform-owners.
Probably people realised that in many cases, a chatbot simply wasn't a good choice of interface for the product/service they were building. Yes, they can work well in some cases, like with help desk type support, sales, general gimmickry on chat rooms and forums, etc.<p>But in a lot of cases, a traditional UI (whether visual or command line focused) is simply far more usable than a bot, and people/companies realised that. As much as it may sound old fashioned, talking to everyone is not necessarily better than using a keyboard/mouse/controller/whatever, and having a bot that has to figure out what you want with far less sophistication than an actual person makes it even worse.<p>The technology isn't there yet for chatbots to be as useful as they need to be, and in many cases, bots and chat interfaces and what not simply aren't a good solution for the problem at hand either.
Chatbots are a feature, not a product. I'm not a customer, but I would say Lemonade seems to have successfully used a chat-style bot as the main feature to onboard new customers. To me, I see context-driven "Google" searches such as WolframAlpha as more valuable than holding a text-style conversation with a bot.
Chatbots were the "next big thing" just like how "Blockchain but not Bitcoin" is supposed to be the next big thing.<p>It's just bunch of people who look at existing phenomenon that's working (people use chat a lot on the phone), and then trying to turn it into something completely different (if people chat a lot, that probably means people want to buy stuff over chat a lot!), powered by dumb money (stupid VCs in case of "chatbots", and stupid ICO crowds in case of "blockchains")<p>The reason these things don't work is because the people who build these things don't truly understand why something is working, because the original product they tried to replicate was not easy to understand.<p>Same reason why it's so hard for dumb electronic companies to copy Apple. They think they know, but all they're copying is the superficial stuff, and they never work.
A part of the job of a call centre operator is negotiation with the client - this is hard to do with a bot. Even more so that this is kind of hidden spec, because companies don't want to admit that the role of the call centre is to make it hard to get something from the company while pretending to enabling it.
Chatbots can only work if they learn to communicate in the millennial language using emojis and acronyms - basically more urban dictionary than dictionary. The older generation, preferred talking to people over searching, or typing a message even if the talking meant they were on hold for 20 mins. Alexa solved for those with a voice based interface, as did google, and people realized that talking to a bot is better than typing. With the voice to text conversion equaling human capabilities, chatbots went down simply to be replaced by a newer and more easy to use technology. Only difference being that they did not get the time to be ubiquitous or become the next big thing as predicted. The evolution of the service, and CX was ultimately slower than what was expected and hence they are replaced/not popular anymore.
Do you really have to ask? This AI merry-go-round has been spinning for decades now. I can't imagine people don't know the drill by now. We've all been riding this around, some of us our entire lives. The people who thought this was going to be the next big thing were just the latest fools to be fooled by this merry-go-round into thinking that AI is once again, just barely out of reach and reachable within a few years. As usual, they are proven wrong again and again because AI is so far from being useful for such tasks like having a conversation with humans. The bottom line is AI isn't there yet and probably won't be the next time the hype rolls around. Or the time after that. Or the time after that ... I should just write a loop here that likely won't break in our lifetimes or possibly anyone's.
Chatbots basically killed a startup I was working with a couple years ago. We had really nice apps, with a realistic exit strategy. Then came "The Pivot", and we were all working on chatbots. Didn't take long for me to leave that company, who went out of business not long after.
I for one would LOVE to try a spoken UI for an email client. My thinking has always been that compromising sacrificing plain english with a more strict/limited set of mnemonic keywords would introduce a learning curve but with a bigger payoff in reducing confusion.<p>Also emails delivered from this sort of interface need some means of indication that they were a spoken format, not typed such that the reader properly understands the context.<p>To me, a verbal UI seems attainable by lowering the bar of what we expect from the AI. Lower the need for AI to interpret your requests from general english to limited keywords. Allow for an adjustable vocabulary for spoken word recognition to accomodate a smaller dictionary with less mistypings. Make it clear the produced content is dictated and not hand-written quality
I called this like 2 years ago. I am still waiting for better apps.<p><a href="http://dangrover.com/blog/2016/04/20/bots-wont-replace-apps.html" rel="nofollow">http://dangrover.com/blog/2016/04/20/bots-wont-replace-apps....</a>
I think everyone has called their bank or cable company at least once or twice and found themselves lost in a Kafkaesque loop of not being able to get done what they needed and have subconsciously written 'bots' off as incapable of helping them perform a task, and/or an executive excuse to put tens of thousands of call center employees out of work. Just because it's on a website and the dialpad is replaced with words doesn't mean it's different.<p>Whether or not anyone wants to admit it, the whole "talking to robots" thing... I mean, the world is already empty and lonely enough without adding another layer of "inexorability" to the processes we engage in, in order to accomplish pretty basic tasks.
The discussions here (at least what I've read so far) have been about customer service bots. That's not where chatbots are going to be fun and interesting. Fun and interesting is the "expert in your ear" that we'll soon have.
I agree with the article in general when it comes to NLP, AI and the dull hype machine. But in certain domains well understood usage patterns in conversational interfaces have advantages over custom UI's. Conversational interfaces with micro UX elements, maps buttons images selectors ..., rather then text input. They advantage is a universal timeline. Be it booking a table in restaurant, ordering a car, searching for an article, all conversational interface allow you to go back in time in a uniform well understood way. By scrolling back. Also your interaction history is searchable. No need to introduce a new UX element for that.
Chatbots were the next big thing before VR was the next big thing before 3D printing was the next big thing before...<p>You get the point.<p>Crypto is the next big thing right now and like all these "big things", it's going to fall flat on its face as well
Chatbots represent an extremely and unnecessarily pessimistic view of humanity. We are interfacing with a machine, that has a certain API. Now, we have two choices:<p>(1) Assume humans are too stupid to be able to adapt their communication and learn something about what that API is capable of and what commands it understands.<p>(2) Credit humans with the ability to do it; perhaps even evolve their communication abilities a little to go along with this brave new world populated by AIs.<p>No intelligent human wants to talk to an API in natural language when they have some idea of the capabilities of that API and the commands it understands.
There's a lot of discussion here about hype, but I think many are overlooking exactly why expectations were so high.<p>Downloads of new mobile applications have been decreasing in growth steadily for years now[0]. Instead, user's spend more time inside their existing applications, particularly messenger applications. Take a look at just a few of their user bases:<p><pre><code> - Facebook Messenger - 1.2B+ [1]
- Whatsapp - 1B+ [2]
- Kik - 300m [3]
</code></pre>
Considering that writing a single bot that was capable to connecting to all of these platforms was trivial, it gave the developer access to a new and very significant user base.<p>Add to this the fact that Facebook, Google, Microsoft, Slack and others, with a combined market cap in the trillions, were simultaneously contributing to the support of chatbots in one way or another.<p>It's a classic rendition of the Hype Cycle[4]. We're probably somewhere in the <i>"Trough of Disillusionment"</i>, but give it a few years and we'll get to the <i>"Plateau of Productivity"</i>, and chatbots will probably make their way to a more useful purpose.<p>[0]: <a href="https://techcrunch.com/2015/09/29/forget-apps-now-the-bots-take-over/" rel="nofollow">https://techcrunch.com/2015/09/29/forget-apps-now-the-bots-t...</a>
[1]: <a href="https://techcrunch.com/2017/04/12/messenger/" rel="nofollow">https://techcrunch.com/2017/04/12/messenger/</a>
[2]: <a href="https://www.theverge.com/2016/2/1/10889534/whats-app-1-billion-users-facebook-mark-zuckerberg" rel="nofollow">https://www.theverge.com/2016/2/1/10889534/whats-app-1-billi...</a>
[3]: <a href="https://techcrunch.com/2016/05/11/kik-already-has-over-6000-bots-reaching-300-million-registered-users/" rel="nofollow">https://techcrunch.com/2016/05/11/kik-already-has-over-6000-...</a>
[4]:<a href="https://en.wikipedia.org/wiki/Hype_cycle#/media/File:Gartner_Hype_Cycle.svg" rel="nofollow">https://en.wikipedia.org/wiki/Hype_cycle#/media/File:Gartner...</a>
With most advanced interfaces or tech, people adapt their interactions to the tool as well. If the tool is capable of evolution, then there is perhaps an equilibrium point where people can be productive with the tool in an idiosyncratic way. This doesn't seem to be considered by people proposing AI agents. "Enactive cognition" - where cognition is compared to the way a cell and its environment mutually define the boundary between each other - is perhaps the model to pay attention to here.
My city went to one to pay for water, trash pickup, etc. It's a horrible experience to use. It takes longer and is less efficient than just logging into a modern website and paying.
Perhaps even more worrying then that nothing came out of it: now all that money has flowed into the hands of short-term opportunistic thinkers who might amplify or incentivize even more hype driven companies in the next iteration of this mechanism. Sure, by no means are all of them bad apples (some just jumped on the bandwagon but will long term perhaps become the next Elon Musk), but I think a strong argument could be made that there are enough people of that ilk in the cohort to be worried.
I have someone to chat to and they're aren't a bot, and maybe they have like a lifetime of experiences from a nation I will never really know and cannot really imagine. It's cool and I have utmost respect for it. I've worked with humans from places where food is considered sacred and given away for free, places where the Volkswagen is almost the only auto to be found, and humans who have survived genocides. They are worth a premium, sirs!!
The two biggest flaws with building chatbots are a) the lack of good tools to express chat flows at a higher level (no, modeling it as a graph isn't good enough) and b) no Natural Language Generation that's good enough to trust in production.<p>The first one leads to subpar bots that can't respond to most things, the second means humans have to fill the 'long-tail' of potential responses by hand, which is impossible.
I think the only way you could think this was "the next big thing" is if you never did M-x doctor in an emacs buffer.<p>LSTMs do reasonable cocktail talk if you want someone to answer your emails for you as an agent, and expert systems shells are actually useful for navigating legal documents and insurance forms (with a ton of human intervention). Otherwise, it ain't really a problem that needs solving.
The people that I know that heavily use chatbots are teens on discord and they use them a lot. They are more versatile than the old bots on irc, use them to play games, stream audio, images, access wikipedia... one bot we use actively translates between different languages across different channels. And they are popular! I think this is a case where the media sees possibility but adoption takes time.
Just like many other here, I feel very smart now :D.<p>My friend was all neck deep into hype. Yet never could explain how would he train the chatbot to learn all the facts about the business(es). And in end all chatbot could do is act like decision tree with some NLP to extent it could understand few variations.<p>Which is great but very few limited business use cases, and poor use case for support unless you want your users to hate you.
Well, any and every promising new thing gets overhyped.<p>(because the press, and experts -- whether real or just self-claimed -- and enthusiasts have plenty of incentives to hype new things and few reasons not to)<p>But that's independent of whether or not the new thing actually has merit.<p>Even when the new thing really <i>is</i> the next big thing, the hype almost invariably runs far ahead of schedule.<p>Personally, I think the jury is still out on chatbots.
Like self driving cars freeing humans from the tedium of driving while they could be doing productive things; applying AI to chatbots could free humans from ever having to chat online. Ever.<p>Imagine how humans could do more useful and productive things. More time for study and contemplation. AI powered chatbots could relieve us from the burden of ever having to chat online.
Airplane is not a bird
Ship not a fish
Car not a horse
....
Computer is not a smarter human
AI will never have human intelligence
Chatbots will never be humans
...
These and many other inventions are useful. They extend human capabilities. But will never be humans, unless we create a human (as opposed to giving birth).
Chat bot was terrible UX to begin with, because while you're at the screen, why not have a visual interface to show you all your options?<p>Voice interfaces hold the promise of releasing us from the tyranny of screen time. I imagine walking the dog and replying to emails with my headphones. Might seem silly now, but would love that. NLP/NLG just isn't there.
I'm surprised how most chatbots focus on a 1:1 interaction with the user and the bot. I find chatbots most productive and novel when you allow the one bot to interact with a group of users. For example, deploy bots, where each user can have the bot deploy as well as they can see the chat history to get context on what has been deployed.
To me, chatbots were nice because:<p>- I could navigate them much faster than I could navigate a phone tree.<p>- They were very accessible in the sense that I wouldn't have to install a custom app but instead use a platform that I am very likely to be on already. As a developer, this also allowed me to set up an unlock-my-door app for my friends much easier.
They're awkward, and they often don't do what you expected or hoped.<p>I think millennials especially are used to getting things done themselves since they're so accustomed to UIs that walk them through getting tasks done. There's a sense of finer control in user interfaces that chatbots have a hard time solving for.
I wonder if, if someone released a study showing that using rainbow colours in your marketing emails increases your perceived humanness by 40%, we'd see a phase where all marketing emails come in bright rainbow colours and I could create a simple filter rule to get rid of them.
I think the big issue really for me personally has been that I know that these interfaces are clunky/inaccurate and I don't want to be talking out loud and giving out info in public spaces etc.<p>If there is an option to type, I got for it the first thing. Anything else feels gimmicky.
There just aren't compelling use cases yet, though eventually the technology will improve and they'll resurface. This happens frequently in tech.<p>On a similar note, I suspect in another few years we'll be reading an article titled:<p>> "AR was the next big thing: what happened?"
Chatbots were always a technology looking for a problem to solve and lets face it those of us reading HN knows exactly how that always turns out. There will be lots of chat bots in our future and already are but there is nothing "big" about it.
Any time a HN discussion gets as big as this, it's because he content is controversial and people are arguing.<p>I don't think I've seen this many replies in unison. It's an amazing testament to how poorly these systems work at all levels.
This is crazy, people actually believed chatbots were going to catch on? My employer developed chat bots, but only so we could grab the associated PR from a few press releases to ride the hype. We knew no one was going to use them.
For the longest time chat bots were a nuance, if you didn't need it it was there and if you did need it, it was there and it would take all day. Those unnecessary pauses to make it seem more human is such a waste of time.
In perfect timing, Bank of America released Erica, "your virtual financial assistant."<p><a href="https://promo.bankofamerica.com/erica/" rel="nofollow">https://promo.bankofamerica.com/erica/</a>
The AI winter is upon us and its first casualty is the chat bot.<p>After all, if AI is as ubiquitous and threatening as many are claiming, then why has the literal least complex implementation of it, a chat bot, failed completely?
This article sets chatbots as endangered, and then works very hard to reset their context.<p>I'm don't work within an industry where chatbots were a solution.<p>From this naive position, what are the exemplary examples of chatbots?
Yet another tech that nobody wants. They are the next worst thing after the "allow notifications" pop-up that every landing page seems to be infested with these days.
Chatbots, like most tech buzzwords, were always a scam. The only question was if you (investor/exec/“researcher”) were dumb enough to fall for it.
conversation ui developer here and I know I might be biased.
For sure there is a hype. People seems to think that chatbot/voice assistant will be the new platform. But really chatbot/voice assistant are meant to augment the current platforms.<p>We can already see this with voice assistant integrated to mobile devices and in the near future VR's and robots.<p>It will be interesting how it works out once everything gets standardized.
They flunked the Turing test.<p>In a serious vein, they were more of an annoyance than help.<p>Similar to the Paper Clip from Microsoft. (Sorry, don't remember the name)
every 2-3 years a certain buzzword gets hot among silicon valley circles to be on its way to become the next big thing. usually after it mostly dies down in less than a year another buzzword will be propelled to the top. thats how silicon valley works, thats how they keep investor money flowing into the industry.
Chatbots, just like Alexa, OK Google, and Siri, are just terrible. Last night, "Alexa, show me new releases." Nothing. "Alexa, show me new movies." Nothing. "Alexa, what's new." "Today's news from NPR....". Damnit. In the time I spent trying to discover the hidden menu of voice control, I could have manually gone through known menus including typing with arrow keys.
David Cancel will probably still sell yet another company for 7+ figures eg <a href="https://www.drift.com/" rel="nofollow">https://www.drift.com/</a>
The best one I have seen is going to start beta testing later this summer.<p><a href="https://m.youtube.com/watch?v=bd1mEm2Fy08" rel="nofollow">https://m.youtube.com/watch?v=bd1mEm2Fy08</a>
Google Duplex Demo from Google IO 2018 - YouTube
On the other hand, I can think of at least two prominent elected officials (one at my state's level and one at the national level) who no longer speak in sentences that I can parse with my own mental model of English. The one from my state released a statement yesterday that seemed identical to the kind of thing one would get from a late '90s IRC chatbot that was trained on a big dataset of energy company press releases.