Two decades ago, we watched movies such as I-Robot,Two decades ago, we watched movies such as I-Robot, and A.I. and wondered what it would be like to coexist with non-organic intelligence.
Now we know.
Machine Learning or ML is everywhere.
Alexa can dial your mother, tell your five-year-old a bedtime story, and wake you up for a morning podcast.
Facebook’s Face Recognition software got so good that they had to shut it down last November.
But the biggest impact has been on business and industry. There has been a transformation of manufacturing through artificial intelligence.
Machine Learning, a subset of artificial intelligence, can crunch large data pools and use it for predictive purposes. Instead of replacing humans, the standard subtext of most dystopian sci-fi movies, ML has become an all-important adjunct that we can’t do without anymore.
Which sectors are most influenced? Read on to learn more.
Application of ML in five business sectors
The application of Machine Learning is really being applied in different sectors and fields. Few of such sectors are given below.
-
Finance
Because of obvious reasons, ML and the finance industry are well suited. ML loves data, and the financial sector churns it out like there is no tomorrow.
A bank generates a vast number of data points about its customers. Not only does the bank know how much a customer earns and spends, but by tracking cards and points-of-sales, it knows where and at what time expenses happen.
A decade ago, credit score was an essential criterion for giving loans. Machine Learning software can find better correlations than traditional income and expense patterns.
Stock brokerages too depend on ML extensively.
Days spent hunched over candlestick charts to figure out the bottom and the top are over.
The guesswork has been taken over by software that has analyzed every stock market movement in the past hundred years and factored in not a dozen but a hundred variables.
Danelfin has developed an algorithm that checks 900 data points daily to analyze 1000 US and 600 European stocks.
The platform calculates which stocks will rise in 30-90 sessions. Goldman is now hiring coders as traders so that their trading floor personnel are at home with tweaking ML algorithms.
-
Healthcare
There can be no doubters for this one.
Machine Learning and predictive software played an enormous role in developing Covid-19 vaccines.
What would have taken five years was done in less than a few months.
Najat Khan, chief data science officer at Janssen Research & Development (acquired by Johnson & Johnson), and Dimitris Bertsimas of MIT Sloan School of Management, put together an ML algorithm named DELPHI speeded up research.
There is no need to go so far.
The hospital near you already has or will soon equip itself with ML-driven treatment plans. That is achievable with ML scouring the data of millions of patients and finding the most effective solution given a patient’s history and comorbidities.
The result? Expert doctors can focus on patients with greater distress while interns armed with ML-based treatment software look after the less severe ones.
The same is true for molecule discovery.
Massachusetts-based Generate Biomedicines uses its ML platform Generative Biology to sift through hundreds of molecules in a matter of days.
Turbine, a biotech firm from Budapest, blends computer science with molecular biology to create drugs for cancer.
-
Insurance
Insurance products are expensive, yet they are not tailored for specific customers, which has been the industry’s biggest drawback.
There has to be a correlation between prescriptions and lifespan. Till now, that type of analysis has not been possible.
Now that is being rectified, and ML helps actuarial science use predictive models besides bare statistical analysis.
Progress is slow since the industry is run the old-fashioned way and is highly regulated.
However, a McKinsey study has identified potential in the following areas.
- ML can be used to price insurance accurately. Now premium charts use the broadest clusters, such as age. But that alone is hardly a predictor of health and longevity, and life insurance and health insurance would benefit greatly.
- Insurance in the US alone is subject to $40 billion worth of fraud, which drives up the premium for everyone else. ML is excellent at recognizing patterns and red flags potential frauds in advance.
- Using chatbots is ideal for answering the bulk of customer care. The resources set free can be deployed for better use in other departments.
Firms such as Accolade are at the forefront of this change. Their Maya Intelligence platform helps patients find the most cost-effective health care plan. With a million happy clients, the system has proved its worth.
Indian fitness wearables startup GOQii Health is partnering with Max Bupa Health to track fitness parameters.
Of course, some of these measures come dangerously close to an invasion of privacy, and regulatory oversight might be necessary.
-
Marketing
This is the most visible aspect of ML.
Open Amazon, and a dozen products are suggested.
Open Facebook offers to buy the best Darjeeling tea or Colombian coffee at a discount.
Open YouTube, and a meditation music track await to put you in the right mood.
The internet brought digital marketing. The term comprises all avenues of marketing that are delivered through digital channels, e.g., social media and search.
Whatever one does on the net can be tracked very easily. The amount of time spent on a site, type of queries asked, bounce rate can paint an accurate picture.
Google and Facebook have an excellent idea of who likes what. They also know where a person is at any moment (correct to ten feet), thanks to the ubiquitous smartphone in their pocket.
Using this personal information with metadata from 2 billion humans to suggest products is just one step.
So successful is it that small businesses with a meager advertising budget can match the reach of behemoths a decade ago.
An entirely new industry, known as digital marketing, has sprung up, employing millions in every country.
Of course, the industry is more than software alone, but the backbone rests on a lot of data crunching by ML programs.
-
Automotive
We have all solved reCaptcha on Google. Find out all the hills or bicycles and click verify, which commands the God of Mountain View, California.
We meekly obey since “suspicious account activity” has been noted.
But you are unwittingly training Google AI models. The images of trains and planes are to let a driverless car learn the difference between the road and a cyclist.
Not only are cars learning how to drive through ML but at a lower scale using driver assistance.
From adaptive cruise control to automatic brakes based on sensors ML is making cars safer for motorists and pedestrians alike.
What does the future hold?
To say Machine Learning is revolutionary is to state the obvious.
It has changed many sectors till now, but it would change the way business is done over the next decade.
From over the next decade manufacturing to hospitality, there is a scope for ML to make things better.
Why is ML so effective at promoting positive change? It is because it tries to read the future based on the past. There have been attempts to do that previously too, like astrology.
ML is a lot more dependable and complex than peering into the night sky and telling about tomorrow.
New applications of Machine Learning would make it an indispensable part of future business models.
Sources
https://www.businessnewsdaily.com/9402-artificial-intelligence-business-trends.html
https://about.fb.com/news/2021/11/update-on-use-of-face-recognition/
https://www.kiplinger.com/investing/stocks/604067/can-ai-beat-the-market-10-stocks-to-watch
https://unfoldlabs.medium.com/the-impact-of-artificial-intelligence-in-healthcare-4bc657f129f5
https://news.mit.edu/2021/behind-covid-19-vaccine-development-0518
https://www.pharmaceutical-technology.com/features/new-generation-ai-drug-discovery-companies/
https://www.nivabupa.com/press-release/max-bupa-inks-alliance-with-goqii-and-swiss-re.html
https://towardsdatascience.com/are-you-unwittingly-helping-to-train-googles-ai-models-f318dea53aee
https://www.eenewsautomotive.com/en/adaptive-cruise-control-uses-ai-to-learn-drivers-behavior/
