Practical Artificial Intelligence

I have been reading a lot about Artificial Intelligence lately. In the years ahead I believe that the prevalence of AI is only going to increase, changing the way we live and affecting every person on this planet. I'll be writing a few posts on AI in the coming weeks as it is an area that personally interests me, as well as being increasingly central to the area of technology that I focus on, the Microsoft Power Platform.

AI seems to be the favourite word of marketers at the moment, whether they’re selling AI toothbrushes or adding AI to reporting and analytics tools to try and make them sexier.

In overuse, though, the meaning becomes lost. The power of Artificial Intelligence comes from giving humans the ability to enhance their existing cognitive abilities at scale, as well as learning from all data to create better outcomes.

Some people worry that it's going to take over the world or create mass job losses.

I think what we will actually see is that as AI is applied in more contexts it will create opportunities for people to do more meaningful work. That will likely mean a change in the skills that people will need. Rather than sticking our heads in the sand we need to realise that the skills of yesterday will not be the skills of tomorrow and we need to develop our learning muscles, with daily learning of skills of the future.

“artificial intelligence can and should automate the regular and the mundane”

You might be surprised about some of the ways AI is already integrated in our day to day lives and the way we do business.

Artificial intelligence is not new, it's just that computing power has now moved into an arena where it is possible to have algorithms process information at scale.

With AI it’s possible to combine satellite and topographical data sets with customer data sets. Allowing, for example, household insurers to provide tailored insurance policies based on an individual's needs and location rather a one size fits all policy.

AI is already being used by some governments to manage taxation. An algorithm can find patterns across their data sets, taking into account socio-economic and role-level factors to determine expected taxation levels for an individual or organisation. It can also spot anomalies and be used in the assessment process to determine their tax profile.

As long ago as 2008 Google has even able to use AI to predict when there is going to be an outbreak of flu. Imagine how many people would need to present with flu symptoms at GPs and hospitals before they had enough cases to identify an outbreak. By monitoring thousands of searches over a short space of time AI is able to more quickly and accurately spot a trend.

You can already start modeling some very sophisticated scenarios with AI, but we’re only just getting started.

Once you have been able to train an AI system with a level of sophistication, repetitive work goes away. Based on existing technology, it doesn’t seem like a stretch to imagine garbage collections will soon be done by robots instead of people. Let’s stretch that scenario. In the future, AI enabled machines won’t only empty rubbish bins, they’ll keep our streets in pristine, clean order. Automatically scanning for any trash that is out of place, sorting rubbish for recycling as it is collected, perhaps even reporting the need for broken bins to be replaced. While these aware machines are on the street anyway, why stop at trash?

The same machines could be empowered with council data sets and be able to match the number plates of parked cars to identify stolen vehicles, as well as broken street signs, or keep an eye on the nocturnal activities among the local foxes.

Your experience at the McDonald’s drive-thru will be smoothly managed by AI. In this example from 2016, you can see that using AI will remove the human element that can lead to mistakes, and even handle a sophisticated, multi-part order. AI is able to understand the context of each order, the specific elements – no pickles on that, extra mayo on this, hold the fries on that. And can use a leveled-up intelligence that allows it to accurately capture the order.

You might argue that we need slower food, not more fast food. But there’s no reason why this technology couldn’t make ordering in any industry much more efficient.

Check out this example from a few years ago.

https://youtu.be/Y_rSb4Vb5zA

AI is already driving advances in medicine. In the future, I can see surgical risks being reduced even further thanks to AI. Say you need your appendix removed. Imagine if you could be operated on by a robot that has looked at, and learned from, every diagnosis and every appendectomy that’s been done. The margin of error goes way down, and the machine would be able to be far more accurate than a human hand. Now apply that to cancer diagnosis and tumour removal.

I think we're moving into an era of precision and accuracy, and no distractions, which will potentially erode even traditionally high skilled work.

Take lawyers as another example, a profession that relies on people learning and understanding huge volumes of data. Stacks of law documentation, legal precedents, and case-specific research. In the future, an algorithm could be trained to read and understand all those data points then potentially make a risk-based analysis or work the odds around to build a case.

I think anything that requires access to massive amounts of data, to make an accurate decision is going to lend itself nicely to AI in the future.

I can't wait till we have AI combined with nanobots so that while I sleep at night, my nanobot will shave my hair, wherever it's located, to the exact length that I like. My beard will be trimmed perfectly, my bald head will be super smooth when I wake up in the morning. I’d love a world where I don’t have to shave every day.

How about a health nanobot that would be constantly cleaning my arteries, removing any type of plaque build-up, whether it be in my brain, to prevent Alzheimer's or any part of the body, to keep everything in optimal condition. Always looking for the early signs or warnings of potential health issues and addressing those directly.

When it comes to AI, I don't see that ultimately it will be one super-intelligent being. I don't think that the same AI that calculates my tax, would also be plugged into my toothbrush and give me the best, cleanest brushing experience and assess the health of my mouth.

I do see millions of interconnected AIs or algorithms trained to do one thing better than a person. Whereas a person is multi-skilled or multi-faceted, AI systems can be built to do one thing extremely well.

Then the power of that AI comes together when you combine these highly skilled AIs with a network or the internet. And therefore, multiple AIs could work together to create a higher degree of value.

What I'm searching for right now is examples of how we can practically apply AI today. Whether it's in health like uncovering disease prevention, or even cures, or how AI can be used to correct humans potential failings of the past (e.g. climate change).

If this are interests you too, I'd love your help answering:

How does AI apply in our everyday lives and make it better?

How can we automate more of the repetitive mundane of our lives?

I am still yet to see a high degree of practical AI. But I feel like as I'm becoming more aware of it, more examples will start cropping up. I think the future of AI is incredible, I look forward to a time when it transforms every part of my life.

I know that some people find that prospect daunting, especially concerning the erosion of privacy. I think that might be the price of change. But what do you think?

I am excited that through the Microsoft Power Platform, Common Data Services (CDS), and Azure Cognitive Services Microsoft are creating practical tools for business to realise the positive potential of AI today.

I think the future is bright – I think it's exciting where things are going. But it highlights the need for everyone to keep up-skilling if we want to avoid becoming irrelevant.

Let me know in the comments below what your thoughts are on AI and whether it's going to change the world.