There is no AI without IA

by May 27, 2020 0 comments

Artificial Intelligence requires Machine Learning; Machine Learning requires analytics; and analytics requires the right data and Information Architecture (IA). In other words, there is no AI without IA, explains Anil Bhasker, Business Unit Leader, Analytics Platform, IBM India/South Asia.

Is Big Data and AI proving useful to track the progress of the COVID-19 pandemic? How is it being leveraged?

The pandemic having spread all over the world has generated data in various forms, at different rates and high volumes—which has spurred the need and use of Big Data analytics and AI in predicting the spread. Big Data analytics helps all this data to be properly transformed into information that can then be mined for purposes like researching the spread, creating prediction models, etc. The flattening of the curve that we hear constantly comes from having analyzed the data.

At IBM, we have been helping government agencies, healthcare organizations and academic institutions throughout the world use AI to put critical data and information into the hands of their citizens. The AI-enhanced interactive ‘Incidents Map’ of COVID-19 data & stats in IBM’s The Weather Channel app & is helping our users stay informed of the ongoing pandemic. The Indian Council for Medical Research (ICMR) has collaborated with IBM to implement a Watson virtual agent (called Watson Assistant) on its portal to respond to specific queries of front line staff and data entry operators from various testing and diagnostic facilities across the country on COVID-19. This is helping boost ICMR’s response time and is allowing them to concentrate on priorities like developing & updating testing & treatment protocols and guidance for COVID-19.

In the future, how can we use AI to monitor and predict the rise of pandemics?

Analytics and AI does a great job of finding patterns & trends in data; and with the proper modeling, data from unstructured (news, social media, etc) and structured (patient data, etc.) can be used to predict spread and also prescribe proper approach to minimize it. AI is very useful in the medical industry, but there are a lot of regulation and privacy issues.

Do you see a forward movement of AI in this field thanks to the urgency that has resulted in the current pandemic? How can AI help here?

This pandemic is a wakeup call for everyone—this shows how fragile our connected world is and we need to ensure to leverage the positives of a connected world (the availability of data) and funnel it properly towards analytics and AI to ensure positive outcomes. Having said that, privacy is a very valid and important factor; masking cleansing the data so only the pertinent data points that don’t single out a person can be used in the modeling. IBM Watson Health was created to help solve some of the world’s most pressing health challenges through data, analytics, AI and hybrid cloud. IBM Watson Health helps health professionals and researchers around the world translate data and knowledge into insights to help them make more informed decisions about care for their patients. It has delivered unprecedented insights—trusted, secure and actionable information we could also use to train Watson in value-based payment models, radiology, oncology and clinical trials.

In today’s Work From Home era, digital transformation will be speeded up. How can organizations big and small use AI to and transform and streamline their businesses?

The pandemic has created a paradigm shift in terms of our perception of business as usual. There are many ways AI can help—helping employees find information quickly and work efficiently, build more secure connections, ensure proper work-life balances, etc. To break it down, AI requires Machine Learning; Machine Learning requires analytics; and analytics requires the right data and Information Architecture (IA). In other words, there is no AI without IA. These capabilities form the solid rungs of what we call the AI Ladder—the increasing levels of analytic sophistication that lead to, and buttress, a thriving AI environment.

We also recently announced a broad range of new AI-powered capabilities and services that are designed to help CIOs automate their IT infrastructures to be more resilient to future disruptions and to help reduce costs. Watson AIOps enables organizations to introduce automation at the infrastructure level and is designed to help CIOs better predict and shape future outcomes, focus resources on higher-value work and build more responsive and intelligent networks that can stay up and running longer. Accelerator for Application Modernization with AI is another new capability designed to help clients reduce the overall effort and costs associated with application modernization.

Have any of your research projects helped organizations during the lockdown and how do you see the future of AI research going from here? Will we become more dependent on AI?

In our efforts to fight the COVID-19 pandemic at a global level, IBM, in collaboration with the US Department of Energy launched the COVID-19 High Performance Computing Consortium—that brought together Government, Industry and Academia to fight coronavirus with all the tools including some of the sophisticated super computers ever built. This consortium has made available 400 petaflops of computing power (for free) to researchers and scientists everywhere, accelerating their efforts to understand COVID-19, its treatment and potential cures. The consortium recently approved NIT Warangal to run an experiment on the study of dependence of structure and dynamics of novel SARS-CoV-2 on temperature and humidity in the atmosphere.

At IBM globally, to help researchers access structured and unstructured data quickly, IBM Research has developed a cloud-based AI research service that has ingested a corpus of thousands of papers from the COVID-19 Open Research Dataset (CORD-19) and licensed databases from DrugBank, and GenBank. This tool uses advanced AI, allowing users to make specific queries to the collections of papers and extract critical COVID-19 knowledge—including embedded text, tables and figures.

We have also granted free access to our considerable patent portfolio to those developing technologies to help diagnose, prevent, contain or treat coronaviruses, including the one that causes COVID-19. Our pledge covers thousands of IBM AI patents, including Watson technology patents, as well as dozens of active U.S. patents in the general area of biological viruses.

Furthermore, to help researchers generate potential new drug candidates for COVID-19, IBM has applied our novel AI generative frameworks to three COVID-19 targets and have generated 3000 novel molecules. We are sharing those molecules under a Creative Commons, with the hope that the research and drug design communities can accelerate the process of identifying promising new drug candidates for coronavirus and potential similar, new outbreaks.

What about the dangers and issues of bias in AI? There are still a lot of apprehensions among the common populace, thanks to things like rogue AIs in Hollywood movies.

Hollywood’s version of AI is generally too dramatic and futuristic. But, bias is something that everyone needs to be concerned about. The need in the AI space is to be able to bring trust in the models. IBM has done a lot of research in this space and built solutions that helps data scientists build unbiased AI models and also help enterprises monitor & de-bias the models (in case bias is detected) without having to retrain the model, thus saving time and effort. Another aspect that helps build trust is explainability—IBM has built the capability that is able to say in plain English on why a model predicted a particular outcome. Another key aspect that I touched upon earlier is IA, which has to be robust for any AI to perform unbiased. If the right & trustworthy data is not fed to the downstream AI application, the output will be skewed and the very purpose will be defeated.

No Comments so far

Jump into a conversation

No Comments Yet!

You can be the one to start a conversation.