As I'm talking with people about my project I'm trying out a great number of different approaches, to try to find what makes it the easiest to understand for most people. So I thought it fun to over time perhaps share a few more of these old and new approaches; like this one (which is an unfinished one not shared before):
Let's say that you're a local startup mentor/investor keeping track of a great number of local businesses.
Every day you (ideally) spend a couple of hours looking at some of these businesses' websites to stay on top of how they perform as far as updating web stores, blog posts, and so on.
To speed up this process you could take your list of web addresses and go to an app developer that make you a unique social network-like app in which you see all these website/-shop updates, in a similar way as you would see posts on Twitter, Instagram, or Facebook. Your own app to give you a quicker way to review all updates, as well as being able to set up notifications for some of the updates.
However, you could not have an app developer build you that app without giving him your list of web addresses.
That is because there's no way to in code do things like: 'from("example.com").get("blog")'.
Extracting information from a website, especially when there's no room for error (such as when looking up a phone number, opening hours, or physical address), actually is impossibly hard to do without a manual, hands-on, process for each and every website.
When we casually surf the web our brains are automatically doing a lot very complex pattern and context analysis for us to understand what we're looking at; and computers don't even come close to being able to do the same. (AI/ML etc is nowhere near being able to provide a general solution to this problem.)
This means that for you to get that unique app, allowing you to stay on top of all the businesses' updates, your app developer would have to manually check the code within each and every business' website, and then write his own unique code for how to extract the data to process and make it available to you.
This hard technical limitation is very integral to how the internet has evolved.
The example of an investor with a list of websites is highly specific, and a bit forced, but we have a parallel in how most of us today "follow" businesses in our social media apps. So in a very generalized sense we are all that investor in how we want to keep track of the updates from some sources; and this inherent technical limitation is what's forcing us to do that through these third party social network's manipulated feeds.
An example from my own life would be that as I try to keep a vegan diet I want to be updated about changes to the vegan menus at especially my local restaurants.
For this to happen today the restaurants would have to not only update their menus, but also post about that on social media, where I would have to follow them (even though most of their updates would be about their non-vegan food), and where I would have to be a bit lucky to actually get that specific update relevant to me (as the so called reach of a business' social media posts' often can be as low as 10% of their followers).
That's just not practical, especially not with all the noise/information overflow if I were to follow any and all businesses just to get a few rare relevant posts.
Without this technical limitation I could by now instead simply have told my digital assistant (such as Siri or Alexa) something like "let me know when this website changes their vegan menu".
Not only that, but I could also have used different types of qualifiers; such as me only wanting those updates when I'm geographically within a certain distance, or at specific times during the day. All in all allowing me to get a summary "vegan update" from lots of restaurants as I enter different areas/cities/countries.
Personally I would probably primarily use this to simplify sticking to a diet, but it could also be about craft beers, sales of favorite brands, police updates, traffic reports, bird watching communities, friends' blogs, influencer updates, travel shows; or anything else.
But, there's this underlying technical limitation that's keeping us from having apps/Siri-functionality like this. And that's what I'm working on changing.
Dear salespeople and recruiters: No, just no!
But, let's play a game: Guess what I value my daily rate at and transfer your contact details with 10% of that amount (through Revolut ↑), and I'll give you up to an hour of my time in a video call; as long as you don't lowball. 😁
Let’s address some feedback I got from that earlier post about my project…
First of all, that last thing about me “thinking” that I’ve got something good going on is a case of “Nordic humility”. The project came out of about a decade’s worth of fundamental research; and is absolutely solid, both technically and feasibility of implementation strategies.
A short recap of an important core-part of the tech: It makes it possible for computers to live work with multi-source information in ways that today either aren’t possible, or not practical (as in it not passing any reasonable cost-benefit analysis).
And the questions I got were primarily about (1) how an implementation might look from a user perspective, and (2) what a business case might be. I’ll get to both by expanding on my previous example of digital information kiosks/signs.
From a consumer perspective this tech could be used in information kiosks/signs (such as at malls, transportation hubs, hypermarkets, museums, and so on), resulting in much “smarter” content/results; fundamentally due to such signs with this tech more easily can work with live information from hundreds of sources, straight from the authoritative sources themselves. (No stale data, and no third-party services manipulating the results to promote ads or “sponsored” content.)
Simply put: The signs have lots and lots of more data to work with; and it’s always the most up-to-date data.
Long-term this means that a user could not only look at/use such information kiosks, but also simply could touch their phones to a small information point (or just scan a QR-code) to have all this information (and live-searchability) available in digital assistants such as Siri and Alexa. So with your phone touching/scanning something by the entrance to a mall it can start answering any question you might have about stores and products, and even rearrange a digital shopping list based on what stores have what.
Short-term almost the same would be possible, but instead of a deep system-wide integration it would simply open a webpage/PWA, via which the same things could be possible (and such a webpage could of course optionally launch a more specialised native app, if the user already has one installed).
So a consumer perspective is that you could contactlessly scan your phone somewhere, and get this one-page smart search/guide to a huge collection of relevant information; which might be anything at a mall, an airport, a branded neighbourhood, a city, a nature reserve, or anything else.
And a simple business case, that would of course be to be able to provide such solutions; with the HUGE unique competitive advantage of being able to quicker and cheaper build more advanced solutions than what’s available today.
In a more “traditional” current-day type of way I could also essentially add a layer on top of this tech to provide a client with a simple unified API to what is hundreds of otherwise differently formated sources of information. Practically nearly eliminating the costly need for backend developers in many startups/projects. Getting some projects market-ready in weeks rather than perhaps more than a year.
And… I would call that the tip of the iceberg of possibilities.
Today I got my EU Digital COVID Certificate (2/2 vaccinated; Comirnaty/Pfizer–BioNTech).
A short intro/teaser to what my main project is all about…
My project is technically fairly simple, but conceptually it can be a bit of a brain-twister; which is due to it solving a technical limitation that feels like it shouldn't even be able to exist in a world where we know that all data is constantly tracked, collected, processed, and so on.
The easiest way to explain it is to start with a simplified problem:
Imagine wanting to call your local customer service at the H&M clothing company. If you know that their domain name is "hm.com" you can go to their website, where you without too much trouble will find their contact information. So how come that you can't just type "hm.com" into your voice calling-app, and have it automatically grab the correct phone number from the hm.com website?
Well, that's actually a super hard problem (for a computer).
Our brains are awesome at completely automatically analyzing patterns and context; and without that computers can't easily traverse the web to cherry-pick the information that we need.
And the implications of that are HUGE.
Just think about how we today are using "@businessname" to follow businesses through third-party social networks, and imagine what would happen if we instead could follow any information from a business directly from their website (i.e. "example.com"). There would be no third-party manipulating what we get, and we as information-consumers get to choose ourselves if we want to follow their blog, their photos, their offers, their supply/webstore, and so on, or perhaps only their opening hours. Much greater information efficiency becomes possible, when/if we want it.
So practically there has been a very real, and very hard, technical limitation that has kept us from getting social media-like apps that allow us to "follow" any source, rather than just the posts put into, and filtered through, third party, commercial, social networks.
And it's the same with information kiosks/signs; they simply can't live access all relevant sources of information.
For instance, such a sign at a big mall cannot be a thin and easily implemented layer that answer questions with live data from all the individual stores' own websites. So they can't show the customers a compilation of all the stores having exactly the products they are looking for. The information just isn't easily available like that; even though it manually is, to a human.
I'm fixing things like that. And, no, I'm not using AI. It's a new approach, naturally completely open and decentralized; and based on technology simple enough that it could have been part of the web when it launched three decades ago.
Sometimes it just takes the right perspective, the right idea, at the right time, to create some magic; and that's what I think I've got going on here.
The world isn't quite at post-COVID yet, but as we're somewhat approaching some sort of return to normality I'll be updating this website.
In the meantime…
As usual I'm available to consult within the areas of technology strategy and business devevelopment; especially supporting startups and SMEs as a tech expert.
I'm also starting to look at returning to some international work again, but right now that's kept "by request". (Due to recent developments in Hong Kong I will not be returning there, nor to China; which also includes stopovers. Taiwan is ok for now.)
Also, I'm looking at putting together some simple to use tech. products/packages helping small B&M businesses do some COVID-adaptations to how they work. (Currently getting feedback from local businesses; and if anything comes from that it'll be announced here.)
Sometimes you forget important details when 3D-printing; and have to apply a bit of brute force. 😆
This is of course why we always do about a gazillion trial runs; because there's always something that has to be adjusted as a design goes from concept to production.
Verifying the quality of the 3D-printer, that we'll be using to print the "cases" for the hardware I'm designing for a project. (Currently using stock RPI0, as can be seen in the second picture.)
📍(🦠🔐): Uppsala (🇸🇪)