HONING HEALTH ALERTS
Prior to the pandemic, you’d be forgiven for wondering what on earth “WHO” stood for. Now, not a day goes by without some reference to the World Health Organisation. Corporate companies spend thousands every year measuring brand awareness… Did so-and-so message cut-through? Is our logo instantly recognisable? How did the new colour scheme go down?
I expect many in brand marketing would aspire to reach the level of recognition that WHO has accomplished in 2020 but, you know, without the deadly virus aspect(!)
What you still might not know about WHO, however, is they provide a Health Alert service via Facebook Messenger and WhatsApp.
Free to use, the service is designed to answer questions from the public about coronavirus with “prompt, reliable and official information 24 hours a day, worldwide”, whilst also debunking myths about the virus. Following its launch in March, it has been rolled out in 14 languages with more in the pipeline. According to WHO:
“From government leaders to health workers and family and friends, this messaging service provides the latest news and information on coronavirus including details on symptoms and how people can protect themselves and others. It also provides the latest situation reports and numbers in real-time to help government decision-makers protect the health of their populations.
“The service can be accessed by a link that opens a conversation on WhatsApp. Users can simply type “hi”, "hola", "नमस्ते", "oi", “salut”, "ciao" or "مرحبا" to activate the conversation, prompting a menu of options that can help answer their questions about COVID-19.”
Collaborating with the hugely popular platforms of Facebook and Whatsapp seems like an obvious solution. With scope to reach 2 billion people, it’s one of the quickest ways to disseminate vital information to the masses. But “obvious” and “quick” don’t equate to “simple”.
The Health Alert service is still very much a work in progress, with upgraded AI capabilities coming into effect more than five months after launch. These advanced features – developed in collaboration with Sprinklr - were announced by WHO on 28th August: AI-powered conversations and localised COVID-19 information. WHO stated:
“Within days of the AI upgrades, over 100,000 persons accessed the service and sent through over 500,000 messages; they also requested data on countries more than 430,000 times, and logged 27,000 sign-ups for daily alerts. Returning users of the service rose to 80%. To date, the WHO Health Alert has received almost 4 million messages from over 540,000 users worldwide.”
Sprinklr’s pro-bono collaboration with WHO is significant. Proclaiming to be the “world’s leading Customer Experience Management platform”, Sprinklr works with more than 75 of the 100 most valuable brands on the planet including L’Oréal, Microsoft, McDonalds, Samsung and Verizon. Their offering enables enterprises to “market, advertise, research, care, and engage consumers” across 23 social channels, 11 messaging channels, chart, SMS, and email.
In other words, they know their stuff.
For Sprinklr – and other corporate collaborators - to share their know-how for the greater good is a testament to how the pandemic has brought us together at a time when we’re physically further away than ever before. On the other hand, it’s good business sense. The sooner the pandemic is under control, the sooner their 75+ headline clients can get back to what they do best: making money.
Sprinklr added the following AI enhancements, as reported by MarTech Series in late August:
Conversational AI capabilities
The English version of the WHO Health Alert service on Messenger was enhanced with conversational AI capabilities. Users can now type concerns in natural language and obtain official WHO health information, advice, and guidance on COVID-19.
New AI-powered answers
If automatic answers are not available for certain questions, Sprinklr analyses these queries and automatically groups them into themes to help WHO determine which new answers should be added to the WHO Health Alert service. As a result, six new answers have been added over the past month related to questions about antibody testing, catching COVID-19 twice, the end of COVID-19, the end of the lockdown, a second COVID-19 wave, and a recovery period.
The WHO Health Alert now automatically sends images illustrating country-specific COVID-19 statistics. These images will be shared when a user asks for the latest statistics of a particular country.
Daily COVID Statistics Subscription
Users can sign up to receive a daily COVID-19 statistics flashcard from the WHO Health Alert. These alerts have helped increase engagement by six times for the service since April 2020.
But WHO aren’t the only one utilising machine-learning in the fight against COVID-19…
In May, the World Economic Forum explored the role French start-up Clevy.io are playing thanks to their catchily-titled COVID-BOT:
“[Clevy] launched a chatbot to make it easier for people to find official government communications about COVID-19. Powered by real-time information from the French government and the World Health Organization, the chatbot assesses known symptoms and answers questions about government policies.
“With almost 3 million messages sent to-date, this chatbot is able to answer questions on everything from exercise to an evaluation of COVID-19 risks, without further straining the resources of healthcare and government institutions. French cities including Strasbourg, Orléans and Nanterre are using the chatbot to decentralize the distribution of accurate, verified information.”
As with most things in life, however, there are two sides to every story…
In September, the Financial Times published an opinion piece by biologist and co-author of the State of AI Report, Nathan Benaich. Entitled “AI has disappointed on COVID”, Benaich relays how “geeks around the world sprang into action” following WHO’s declaration of a pandemic:
“A company called C3.ai compiled data sets for AI researchers to crunch. The AI-first drug discovery company Recursion released a morphological image data set of Covid-19 infected cells as they react to thousands of drugs. DeepMind used its AlphaFold AI system to predict and publish structures associated with coronavirus.
“AI seemed on track to crush the new pathogen, just as it mastered the ancient Chinese board game Go. And then - nothing.”
The reason? COVID-19 is so new and complex that the data needed to train AI to combat it simply does not exist.
It’s one thing to adapt existing AI-enabled technologies to handle communications at a time of crisis. It’s another thing entirely to expect newly-developed algorithms to win a race against classical biology and chemistry without the necessary data.
What AI needs is time. Under the right conditions - and with enough data - Benaich believes AI can still play a pivotal role in combatting coronavirus, concluding:
“If this is not the first pandemic to be solved by AI, then it will surely be the last to be solved without it.”
A bold statement indeed. Let’s hope he’s right!