2020 Was Impossible To Predict. Can AI Save 2021?
By Andrew Filev
In January 2020, the words “Wuhan” and “Coronavirus” started to appear in global news reports. These still vague terms caused most people to shrug without giving them too much thought. SARS and H1N1, aka “Swine flu” — the much-reported viruses of previous decades — had inured us to the real dangers of pandemics. While those viruses caused tragic loss of life, they never materialized into globally disruptive pandemics — and so most of us had no reason to assume that the newly minted SARS-CoV-2 would amount to anything more than another small-scale health crisis. This time around, businesses carried on, recreational travel continued, and in many parts of the world, governments were slow to recognize that a storm was on the horizon. The point I’m making is that, given the way we humans are hardwired to think categorically, we may not be the best judges of when to panic.
Perhaps, COVID-19 will be the event that scares us straight and the next time reports of a mysterious virus start creeping into the media, the world will come together with a swifter and more coordinated response. (As I’m writing this post, Australia has about 30 thousand total cases nationwide while California has over 2 million, so there’s hope that some countries and regions will learn from others.) If this pandemic isn’t the event that disconnects us from our instinctual, categorical thinking, then we would do well to leverage technology to warn us of impending dangers. That’s why 2021 is the year that businesses and governments should start building an AI strategy so computers can look at data the way humans too frequently don’t. AI can’t yet replace humans, but in partnership with skilled analysts, it can help us compensate for our weaknesses. Here are a few areas to focus on.
Anomaly detection
The strongest and most immediately impactful application of AI is anomaly detection. Anomaly detection is a form of machine learning (ML) in which computers rapidly detect inconsistencies across massive data sets, then flag them to humans for further analysis. Most of us already benefit from this technology in the form of credit card fraud detection or a notification that an unauthorized person has tried to access one of our accounts — but these are just scraping the surface of the capabilities.
The use of anomaly detection AI will likely expand in healthcare to identify concentrations of seemingly unrelated symptoms across geographies, perhaps giving us a jump on locating outbreaks of emerging diseases. On an individual level, AI could monitor data from our wearable devices and monitor our vital signs for impending health risks like strokes or heart attacks. You may be likely to blame your indigestion on something you ate or muscle soreness on overexertion, but your device could help you decide if it’s time to see a doctor.
In a business setting, anomaly detection can help marketers uncover emerging trends in consumer preferences, help manufactures make earlier and safer decisions about product recalls, and allow security professionals to identify potential sniffers before they become active hackers.
Risk detection
Whereas anomaly detection helps find unexpected and unknown occurrences across data sets, risk detection is a more targeted application of AI in which systems are trained based on your known risks and disruptions. They may include supply chain problems, staffing shortages, or other factors that may slow your business down. The volatility of 2020 put this topic in front of business leaders and their analysts, yet many were still caught unprepared and lacked cutting edge methods for contingency and continuity planning.
To make risk detection work, it’s critical to have a large collection of historical data about your business operations that can be used to train the AI. The larger the data set, the smarter your ML will be in making predictions. This would be yet another good reason to implement a system of record for various workflows and projects. Once you accumulate enough data in the system of record to estimate the baselines, the systems can help you notice if certain workstreams get blocked — giving your team the opportunity to solve these challenges before they impact your business.
No-code can bring AI to the masses — if you can keep it fed
Historically, one of the main challenges of AI is that people with the skills required to implement and configure AI systems are in short supply. That’s why no-code will revolutionize AI and democratize its availability to analysts and leadership with a significantly lower barrier to entry (i.e., no Ph.D. required).
The crux of no-code is that like all AI, it still needs data sources to be useful. For this reason, it’s important to stress that AI isn’t a product you buy or a team that you hire: It’s part of a strategy that spans your IT, operations, and analytics organizations and involves integrating systems across all aspects of your business, including work management, resource planning, inventory tracking, revenue recognition, and more. But if you can pull that all off, making no-code AI platforms available to frontline teams can empower them to become smarter and more effective in their planning.
Assume your competitors are leveraging AI
Not every business can flip the switch to an all-encompassing AI strategy overnight, but you can get started today with technology to help you predict common risks like project delays, cost overruns, and identity deviations in your business processes. It’s not a “nice to have,” it’s going to be critical for competition going forward.
In today’s world, where Amazon and other global powerhouses have the ability to enter new markets virtually overnight, you should be operating under the assumption that your competition is already employing AI in planning how to disrupt you. I don’t say that to make you paranoid; I say it to wake you up to reality. If you’re not already investing in AI for your business, the time to do so is now.
"save" - Google News
January 15, 2021 at 05:55AM
https://ift.tt/2LBzv5V
2020 Was Impossible To Predict. Can AI Save 2021? - Forbes
"save" - Google News
https://ift.tt/2SvBSrf
https://ift.tt/2zJxCxA
Bagikan Berita Ini
0 Response to "2020 Was Impossible To Predict. Can AI Save 2021? - Forbes"
Post a Comment