The field of artificial intelligence (A.I.) and machine learning (M.L.) in particular, has developed significantly over the last five years, partially due to advances in hardware and processing power, from companies such as Nvidia, IBM, Microsoft and Google.

We’re now in a position where this hardware, combined with the growing datasets means that companies of all shapes and sizes are in a position where data can be used to train algorithms, before allowing them to make decisions based on information and new data fed into the system. One of the most powerful uses of A.I. lies in its ability to uncover new insights, revealing to us today what was previously invisible. With the 2.5 quintillion bytes of data being created each and every day, there are plenty of these insights waiting to be uncovered.

As mentioned, M.L. is central to this approach as algorithms can identify patterns in the data to help us understand people’s behaviours and preferences. Companies are using feedback loops built into their processes to increase the amount of data being collected, as well as improve the level of accuracy as more and more data becomes available. From marketing, to business operations and finance, these algorithms have significant implications in business from all angles, which is why it it will become a necessity that A.I. takes a pivotal role in all companies in the near future.

One of the most powerful uses of A.I. lies in its ability to uncover new insights, revealing to us today what was previously invisible.

I’m one of the co-founders and CEO of KOMPAS, and we’re using advances in M.L. to analyse and predict movement, behaviour, and the needs of small businesses, to create a seamless city exploration experience. Currently, KOMPAS is part of the world’s leading AI-accelerator, Creative Destruction Lab in Toronto, and we’re working with people all around the world, to use data to drive decisions in our industry, and as a business. Most importantly, you can take everything into account when looking at the data, from the photos that you’re showing a user, to the meta-data and everything else associated with it, meaning that everything you do can lead to a more accurate outcome in the future.

KOMPAS was born out of a problem that my co-founders and I had while living abroad. We struggled to find out what was available in a city, and more importantly, we couldn’t find information that was personalised to our own specific interests. We were never going to have the time, nor money to visit the hundreds if not thousands of venues we were interested in, and all we wanted to find were the hidden, and more unique locations, bars and restaurants dotted around the city. Our passion for technology, data and travel, led to us founding KOMPAS, which is using k-means clustering, and other forms of M.L. to personalise not just content to people and vice versa, but the reviews, the experience and the overall application to both our users, and our customers. As a result, every user that’s part of the KOMPAS ecosystem, helps us to continue to personalise content to new users, while getting an experience that’s different to anything else out there.


Naturally, as the technology begins to understand you and what you like to do, the content becomes more relevant, meaning that you don’t waste time searching for where to go, as you have a personalised guide sitting right in front of you in the form of an app. That said, we don’t want to drive you to the same place over and over again, which is why we’ve built in elements of randomisation, to get you to do things that would be otherwise outside your comfort zone. As mentioned, the feedback loops built into the application allow us to test how accurate these new recommendations are to you, meaning that we can begin to tailor the experience to you if and when you change over time.

It’s this sort of technology that’s going to become increasingly important, not just in mobile applications, and for creating particular use cases, but for businesses and industries across the board, as research has shown that people are placing an increasing importance on personalisation and uniqueness, over the generic and en masse offerings, and we’re at a time in society where this technology is now able to provide a solution, to an extent at least. That said, while there are huge advances in the technology, and there are some key benefits to it, such as those mentioned above in what we do with KOMPAS, there are also some concerns that exist with the development of A.I.

There are concerns around jobs, and the role of robotics and A.I. in the future, but my argument to that, is that new jobs are created as a result, and it frees up time to allow for more creative roles to be established, in fields where A.I. is unable to make huge advances.

There are concerns around jobs, and the role of robotics and A.I. in the future, but my argument to that, is that new jobs are created as a result, and it frees up time to allow for more creative roles to be established, in fields where A.I. is unable to make huge advances. Instead my concerns lie around our increasing reliance on the internet, and the habit that we have where we use the internet and social media to escape reality. I believe that it’s going to be increasingly important that we use A.I. to improve reality, as opposed to drive more people into a virtual world. My second concern, does lie around jobs, but instead it lies around the re-training of people. There is arguably, a limited life-expectancy for jobs such as manual labour, so it’s going to be increasingly important that we as a society are able to understand the role of A.I. in everyday life, and how we work with the technology to create a solution that is beneficial for the many, not for the few. This revolves around social goals, as well as looking at ethical implications of technology. For example, we should not be using A.I. to advance the power of the military through the use of lethal autonomous weaponry, but we should instead be using it to drive effective and more accurate results within manufacturing for example. With all of this being said, I think we’re going to be set for some big changes in general, over the next decade.

For us as a company, the next steps within our M.L. base, include moving towards a neural network, as opposed to clustering, as we’re able to use data to truly personalize each and every experience for a user, that will adapt quickly over time. We want to create a service that’s local, accessible and easy to understand, creating a solution that prevents the number of small businesses failing. With 90{87a18df7a28eb56c6a7dc02e4e1a3d322672f7d5de2b418517971f2bf2603901} of small businesses closing within two years, we’re doing everything that we can to reduce that number using A.I., one percentage point at a time.

Tom Charman, Co-Founder and CEO at KOMPAS

Related:

0 0 votes
Article Rating