Smart fridges, smart phones, DVR’s, the Internet of Things is becoming an important part of our daily lives and everything we do.
Interestingly, our friends at Boldport ironically posted on their twitter account “The Internet of things is someone else’s problem”. This remark got us to thinking… If we, an electronics company, who make and contribute to this industry, do not think about the positive and negatives IoT and AI brings, who exactly will?
At European Circuits, we’re asking; will AI and the Internet of Things progress as expected and will it be a good thing for markets across an almost limitless array of industry? How do we prepare ourselves for the future?
Did you know? In close collaboration with Replicade, we are able to transform companies initial IoT ideas into brand new hardware products!
This topic is obviously a massive undertaking, so we will blog as we go but this month we will give an overview of AI, IoT and Big Data with a view to provoking thoughts, ideas and hopefully our readers will join in on the conversation, sharing their experience of the subject area.
Historically, the world’s view of AI, for the most part, has felt like an other-worldly, futuristic endeavour, in thanks to science-fiction movies and novels.
The term artificial intelligence may lead to frightening connotations for some readers – such as Skynet from the Terminator blockbusters or HAL 9000 from Space Odyssey.
The foundation of AI is machine learning algorithms. The big concept behind machine learning is creating computers that can learn on their own accord without being programmed. We already have it today, albeit on a smaller, narrower scale. Humans have a broad knowledge base, AI has tended to focus knowledge into confined tasks.
Elon Musk references a calculator as a good example of a computer, of sorts, outperforming a human in one specific discipline.
Another mainstream example of machine learning would be Netflix’s recommended section. Netflix give you show recommendations – that’s a form of machine learning on a basic level. The program provides you with new choices to view as it’s able to analyse the shows and movies you watch and determine the themes that interest you.
Incorporating AI into IoT
As big data is still in it’s relative infancy, machine learning is slowly but surely impacting IoT products. Take Tesla as an example, the company offer partially autopiloted car capabilities, and have the ultimate aim of mainstream driverless cars in the next 20 years. The company rely on machine learning from its car sensors. A new Tesla model, brand new from the factory today will have collected all the information that older Tesla models have experienced on the road. Any new variable or experience can then be shared with all current Tesla cars on the road thus making the autopilot mode much safer for us humans.
Think about the market sector you are in and ask could a computer program make it more efficient in the long run – What data would give you the edge?
There is also the flip side to the positives data intelligence brings- how do we control machine learning in such a way that it helps us in our daily tasks but doesn’t make us redundant in the process?
Big Data is the word associated to any amounts of large data collection. The data can be anything, from market research by a marketing firm, medical data collected by health professionals to the miles you run with your training watch on. Information taken from the devices can help industry predict sales changes and purchase choices in the market place before they happen.
An interesting case in point would be the work carried out by Cambridge Analytica, (who’s company mission statement reads “uses data to change audience behaviour”). In the lead up to the Trump presidential election win, the company, reportedly targeted key states with messaging on social networks such as Facebook reconfirming political points or spreading specific news about candidates to people who were likely to spread the news or be swayed by messaging.
Machine learning isn’t perfect and can be subject to human manipulation or subversion. An example being when Microsoft implemented a basic AI to a Twitter account. In less than a day, trolls were able to turn it into a racist mimic.
Regulation & Cyber Security
The consensus among business leaders is that AI, Big Data and IoT needs to be regulated to work long term. If regulated properly, machine learning could make the world a better place, however, if we do not proactively regulate, it may be a case of the horse will have already bolted and we are at the mercy of AI!
An important part of the regulation will surround privacy and safety of networks. Neil (Our Marketing & Business Development) attended a Glasgow IoT Meetup last month and one area that stuck out from the presentations on show was the need for companies to actively research and find a security company to work with at the outset of designing an IOT project.
Security solutions that provide infrastructure assurance, application layer testing, and vulnerability assessments must be an area we all think about at the outset rather than when the product is up and running. (Neil sends his thanks to Paul Ritchie from Glasgow based Secarma for the interesting presentation on the subject).
There is so much more we could talk about, Robotics, Virtual Reality and Economics (Blockchains) being three prominent topics that tend to lend themselves to the futuristic thinking process we are discussing here but this will need to do us for now as the Neil and Philip (Sales Manager) need to get prepared to fly off to Barcelona for the IoT World Congress on 3rd to 5th October. Find out more here.
We’d love to hear your IoT experience and advice too! Get in touch via e-mail or on our social channels.
Looking for a book on the subject? You may be interested in…