There are websites out there delivering everything from essential medical information and government services, to hate speech, fake news and the latest juicy posts from friends about their love lives. We’ve also seen the rise of the bots: bots that intervene to assist us when we’re messaging friends, those that help us with instant chat-support online, and those that annoy us when used to snap up all the tickets for a gig in the first seven seconds.
So, what is the evolution of our interface to the internet - and will it be friend or foe?
The death of the web as we know it?
Websites have been around for decades. First simple pages of information linked together to form an interactive digital booklet or brochure, then increasingly complex creatures that behave more like fully-fledged online applications. Websites now also have quite a bit of intelligence sprinkled around in the form of chatbots that can often feel as real as a human when responding to written enquiries.
But bots can readily exist in other habitats too – they can be found in popular messaging apps (such as Google’s Allo) popping up to help with useful autocorrect and emoji responses or with Siri-like assistance to typed questions, as well as at the end of the phone where a more natural conversation with a robot is often far preferable to the universally-loathed spoken menus.
My favourite poet, Pam Ayres, once wrote about such automated systems in one of her shorter pieces: “If you would like to meet the person who invented this system and shoot him with a gun, press 1.”
Bots, and virtual assistants more broadly, are undoubtedly here to stay and getting cleverer all the time.
Will increasingly intelligent bots eventually do away with the need for websites altogether? Will the labourious job of manually clicking our way around pages of information, navigating through menus and perhaps having to resort to a search or two one day feel as antiquated as writing a message on paper, putting it in an envelope, sticking on a stamp, and walking to the nearest postbox in order to communicate with another person?
Instant information without lifting a finger
In a recent article on Tech City News entitled ‘AI and chatbots: The future of customer service’ Richard Stevenson, CEO of Red Box Recorders, explored a future in which intelligent chat-based interfaces may well completely obfuscate the need for companies to have a website. The bot could be text-based or voice-driven, but the central feature will be its ability to have a natural conversation with the user and be able to answer any question, however complex, relating to its domain.
Once you’ve experienced the ease and satisfaction of asking a virtual assistant (such as Amazon’s Echo or Google Home) to give you the latest news or a piece of information, play music, perform a task such as setting a reminder or alarm, or operating a connected device such as a thermostat or lights, you might never want to lift a finger again.
Inclusive bots = profound benefits
For people who literally can’t lift a finger, or for whom a disability presents other challenges when it comes to using technology, an intelligent bot or virtual assistant that can understand natural language requests has even more profound benefits.
You can get the tiniest of insights into the many ways a virtual assistant like the Echo can be used, by listening to my quick Echo demos podcast (currently at fifty episodes and counting, there's lots to share!)
The vast majority of websites still present huge issues for people with disabilities. Whether it’s unlabelled images, links or buttons that require a mouse, intrusive ads that confuse and distract, autoplaying videos, a cluttered appearance and wordy text, or just poor choices for colours and fonts, the internet is a challenging place.
And just how many of these issues does someone without a disability encounter on a daily basis? Accessibility is certainly not just a disability issue – especially in this mobile-first world.
We’re still some way away from AIs that are intelligent enough to enable us to entirely avoid using websites to manually trawl for information or to carefully complete online forms to use a service or order a product, but we can already do all of those things very successfully in many areas using the existing semi-smart virtual assistants on the market today, and they’re getting smarter all the time.
In a recent episode of iOS Today, hosts Leo Laporte and Megan Morrone demonstrate the relative intelligence of each of the most popular assistants and discuss the pitfalls of trying to provide single definitive responses to questions that haven’t got simple answers.
It’s possible to say how tall the Eiffel Tower is, for example, but even a question as seemingly straightforward as “Can I take knitting needles onto a plane?” could end in big trouble for the individual if the assistant gets it wrong.
Google is trying to overcome this problem in order to give the user a single answer rather than constantly pointing them at a list of search results instead. Many of us have had Siri or Cortana tell us “I don’t know, but I’ve found this on the web for you”.
They are achieving this by employing complex algorithms that compare a number of search results, try to ascertain from them the ‘right’ answer, and then finally provide it to the user while at the same time attributing the response to one of the more reliable sources. So in this case the Google Home AI will respond: “According to the Huffington Post, you can take knitting needles on internal US flights, but not circular thread cutters” (whatever they are).
Called Google snippets, these single answers are by no means always reliable, which is why they attribute the answer to the chosen source.
The natural, flexible, interface of the future
With the continual improvement of bots or virtual assistants that are inexorably helping to turn labourious clicks into natural chats, we may well see the weakening of the web as we know it. An intelligent natural language interface is flexible in the extreme – we are not just talking about cylinders in your kitchen now, but an intuitive interface that could have text or speech as the input or output in either case and with ever-more-sophisticated smarts helping to understand you and swiftly deliver exactly what you want.
There may well be a screen involved (on which the AI could display additional information) along with any number of other helpful output alternatives, but these would all be complementary rather than essential. In this way, the interface of the future would be able to be used as easily over the phone as on your phone and as easily by an eighty year old, as by an eight year old.
Websites are great (when they are accessible at least), but in the future we may well look back at how we access the internet of today and leave us feeling that it must have been, quite literally, like crawling a web.
You might also like: