I participated in the 1st Symposium on Corporate Information and Knowledge Management at the IV TOI - International Congress on Information Technology and Organization that took place on May 17, 2018 at the Brasiliana Guita e José Midlin Library (USP).

Prof. Francisco Carlos Paletta (ECA-USP) of the Research Group Observatory of the Labor Market in Information and Documentation (OMTID) - CNPQ - opened the symposium commenting on the importance of an event like this to share information between the university and the productive sector, also citing the success of the IV TOI with more than 2,000 subscribers.

The first table addressed the impacts of artificial intelligence and data science on intelligence activities. The first lecture was given by Sandro Sinhorigno - Modeling Superintendent of Banco Itaú - and its theme was the use of data to define the release of credit, both formerly and today, with a view to increasing the volume and variety of data, especially unstructured ones. He also cited JPMorgan's COIN (Contract Intelligence) case that automatically analyzes loan agreements, a task that consumed, on average, 360 thousand hours of work per year for lawyers and credit agents. Today, in addition to revising documents in seconds, the software is also less error-prone.

The second lecture given by Prof. Alexandre Del Rey - Founding Partner of I2AI (International Association of Artificial Intelligence) and Engrama - presented the case of Philips, which began to list the data available for free in the Rating (note) and Reviews (comments) of its products on the website of Amazon to create product innovations that really impact consumers and thus increased the sales volume of its Premium line by 15%.

Prof. Andre Saito - President of the Brazilian Society of Knowledge Management - moderator of the panel, highlighted the public's issues related to the profile necessary for the information professional in this current context of big data and artificial intelligence. Mainly, what would be the differential of the human being in this job market, in which computers have a much larger, faster information processing capacity and today with a more accessible value for companies.

The speakers were emphatic in stating that the machine always has a restricted domain, the ability to create links between information, analyze and monetize the data, understand human behavior, know how to ask the right questions, select the really important data, use the data with creativity, they are still exclusively human functions. Artificial intelligence follows the rules and methodology defined by the human being and cannot always analyze different contexts.

At that moment, I remembered a news story I read about a research from the University of Lisbon that is developing a machine learning model capable of identifying messages on Twitter that have a distorted character and thus checking posts that have a tone of irony and sarcasm, something difficult for computers that read information literally.

Sandro Sinhorigno complemented the discussion with the concept of augmented intelligence that is based on systems with cognitive intelligence that supports the human being, his planning and analysis, freeing up the professionals' time for other activities.

The speakers also affirmed the need to have a multidisciplinary team for the success of these activities, since there is hardly a “unicorn” professional who will have all the necessary skills. However, it will be increasingly important that professionals learn the minimum necessary from other areas to be able to dialogue with the other members of this multidisciplinary team.

Both lectures were recorded and are available at the link: https://www.youtube.com/watch?v=LjLj3ff-pvc.

I recommend that all information professionals watch in full!

After the end of this first table, in the break for the next, talking to other librarians present at the event, we discussed how several issues mentioned by the speakers are directly linked to our area. We already have the skills to organize and create links between data, transform unstructured into structured data, locate and select the really important data, ask the right questions at the time of research, we just need to transpose the previous knowledge and methodologies focused on books and documents for this new reality.