Can AI help with uptake of new products and solving the (minor) problems people face?
Can AI help with uptake of new products and solving the (minor) problems people face? Recently I created a phone app (https://www.easyplan.info) and put it on the Apple App Store and Google Play Store. It's free and solves a recognisable problem. I'm sure more than seven people would find it useful. Seven is the number of installs of the app as reported by the two stores. I used the word minor in the opening question as the problem the app solves is hardly earth-shattering, more a niggle in the flow of the day in a handful of quite specific situations. The point is though why is still hard to let the relevant people know this app exists. The internet has provided a massive change in how much we can communicate, how much information can be shared and at very high speed as compared to a time before its arrival. We now have a heavily managed and edited media, in a positive sense, on the word hand and a massive amount of "information" readily accessible on our laptops and on our mobile phones. Some may question if a picture of a cute cat is information but that's beside the point.
The problem of sharing information could be considered as solved with the arrival of the printing press. The problem of access to information could be considered solved (albeit incompletely) by the arrival of the internet, or more precisely, the web. Education enables us to sift through information moderately quickly but it remains nearly impossible to being able to comprehensively and reliably find the information we seek or are even interested in.
We have the fountains of 'information', pouring selected updates at us in Facebook and Twitter. There are some distinctions in colour of fountain - what I read in my feed on LinkedIn has a distinctly different flavour than that in Facebook. There are differences in media form such as Instagram or Imgur being about images as compared to text as found largely on the channels already mentioned. These various channels use algorithms to choose which updates on their systems to bring to my attention - it's fully automated yet based on what someone or, more likely, a group of people decided would be the general rules for deciding what is most likely to be relevant to me, or attract my attention back to the platform at least. Traditional media is heavily managed in that people edit the content and deliver it as a package to me. Some channels related to those traditional ones allow some tailoring, an application of algorithms combined with my selection, to enable me to see what might be more relevant to me.
At the other end of the spectrum is Google (or Bing etc) search but it's still an algorithm combining my request with what other people searching for the same or similar found useful to decide what to show me. Perhaps even further along the spectrum is the traditional physical library. With the physical library there is an objective structure in the Dewey Decimal System layout of the books and there is help from librarians willing and able to help interpret a verbal description to a suggestion of where to look for the information. Nonetheless, someone decides which books to hold at that library.
There are several parts to the topic of acquiring relevant information. One area in this applies is acquiring tools that make my day to day office-based activities easier, more efficient or more pleasant.
The first part is "find" - the physical library and online search tools go a long way to covering that.
The second part is "access" - the physical library may be the British Library that holds every published book (although not everything printed) and the web provides a huge and increasing amount of information useful to many (although still far from everything printed or known).
The third part is "curation" - traditional media does this in spades and digitised versions deal with the previous two parts; the physical library has an organisation system that enables you to go to the relevant category but that might not be enough and not all publications will available at that location; the web has some traditional media available, you can follow what selected people are saying or referencing, many sites have automated recommendation systems promoting content from their systems tailored to you.
The fourth part is "discovery" - working out what the question is and who to trust to answer it. Indexes in books are fairly useful if you know what to look for; the same is the case for Google and you would probably trust the answer found in Wikipedia. Your cry of "there must be an easier way to do x" or despair at having to face up to some task are heard by no-one other than immediate work colleagues and possibly the cute cat on the sofa. In a library serendipity can help you discover the information (or picture or explanation etc) that you seek and therefore gain clarity on your question. Asking someone specifically because of their expertise establishes a solution and it could be money well spent. You might not even now there is a better way or feel inclined to do something about the existing state of affairs even if you despair at immediate task ahead.
Part of it is down to trust, you don't want to improve a situation because that recognises and gives weight to the problem you wish didn't exist. It's even worse if you share that with someone because they then know about it and who knows where that might lead. The objectivity of the physical library or the online search removes that second fear, the first fear is perhaps easier to tackle as a consequence. Paying someone to review a situation and make recommendations relies on getting past those fears and entrusting them to be objective, professional and perhaps even discreet. However, there is a significant cost in time if not financial to engage them so dealing with small things isn't practical down that route. It costs next to nothing to ask Wikipedia for the definition of a "tonne" for example. It's probably not cost effective to ask someone else to find it out for you. They would need to know more about the context and the type and format of answer required for example. In other words, you didn't have the whole question neatly specified in the first place, you just knew what you wanted and went through a short journey, in this case, of discovery.
If I could write down what wanted to know or wanted to improve then could the computer (i.e. a service on the web) help me with the discovery journey? A Google search is a single request-response cycle, I write something and it gives me back links to pages that Google considers most likely to be relevant; there is no journey. Chat bots are back in favour and perhaps will be more successful this time round. For example, Babylon Health's pre-diagnostic app asks questions based on what you write and the stored knowledge of medical information. It is of note that this works in a specific domain of knowledge. Google is for anything by contrast.
So, as ever, it comes down to who will pay to build the discovery tool and who has, or can pay to build, the relevant pool of knowledge. That's hard to do for apps for which there is no charge. Apple and Google take money for posting the apps on their stores and a cut of costs for apps that are not free - they balance matters in such a way that having a huge number of free apps on their stores is possible and practical.
I conclude with this. A piece that is missing is a service which provides an objective journey to discovering tools relevant to me in a fast and efficient way. Perhaps I should build that service with AI tools available today and aim for someone like Apple or Google to buy it eventually. What do you think?