Khipu AI 2019 — Latin American Meeting In Artificial Intelligence

2019 was the first year that this Conference happened in Latin America, as you know AI (Artificial Intelligence) has the potential to dramatically improve the lives of people and increase the prosperity of businesses, but solving a problem as hard as intelligence will require a diversity of thought and the best minds from every corner of the globe. The mission of Khipu AI is to support the advancement of Latin American talent, research, and companies in AI through an annual event. The organizers of this Conference was inspired in Indaba which is the same Conference but organized in Africa.

The application to Khipu AI was opened in January of 2019 for all people in South America and many of latino americans applied to this Conference, but only a few people were chose for getting the ticket. I had the opportunity to attend to this first amazing Conference, thanks to the organizers for let me attend to Khipu AI 2019.

The following are my highlights of Khipu AI 2019:

Lectures

• Lectures were from the basic to advanced concepts. All of the lectures were really educational. I was really impressed about the quantity of the research in medicine, I think we had more of these topic in the conference and speakers talked about how Machine Learning can solve problems that affect humanity like Guillermo Saphiro — Duke Professor with Duke University who is working in a solution with Computer Vision in order to help children with Autism.

• Another thing of the conference that I loved was the participation of women speakers who did interesting lectures like Machine Learning Fundamentals by Luciana Ferrer who is a researcher at the Computer Science Institute, from the National Scientific and Technical Research Council (CONICET) and the University of Buenos Aires (UBA), Argentina. In this lecture, she showed us that good Machine Learning research requires good evaluation practices like supervising training pipeline, evaluation pipeline defining a meaningful metrics. In the next pictures you can see other women in AI with their lectures:

Machine Learning Fundamentals (Luciana Ferrer). Luciana Ferrer is a researcher at the Computer Science Institute, from the National Scientific and Technical Research Council (CONICET) and the University of Buenos Aires (UBA), Argentina

Robotics and Continuous Control (Chelsea Finn). Chelsea Finn is a research scientist at Google AI and a post-doctoral scholar at UC Berkeley

Microdata Anonymization for Learning Analytics (Lorena Etcheverry). Lorena Etcheverry received a BE in Computer Engineering (2003), a M.Sc.
Degree in Computer Science (2010), and a Ph.D. in Computer Science
(2016) from Universidad de la Republica, Uruguay

Workshops

• I loved this part of the Conference because the attendees could apply the lectures in the workshops. My favorite workshop was about Reinforcement Learning were I learned how to create a new model with RL, and the speaker of the lecture (Nando de Freitas) came to the workshop in order to solve any doubt that we had, it was really educational.

Posters

• The quantity of women who attended to the conference was remarkable , and this women not only attended to the conference, they also presented their posters. The posters that I most liked are one of Deep Mind about “Memory in Agents” where they study how well the agent can apply what it learns in training.

Memory in Agents

Efficiently Improved Hierarchical Text with External Knowledge Integration

Networking

• I think the best part of the conference was the Networking because people who came from other countries had the opportunity to be close with the international speakers, this is not usual as in other kind of conferences. Then, I think this was the most valuable experience in this event. For example, I had the opportunity to had a lunch with Nando de Freitas — Principal Scientist at DeepMind and the Bolivian delegation, the conversation was focus on how Bolivia can improve the Education in AI and how tech communities can help to reach this objective. In addition, he gave us a strategy to teach children in AI because they are the future for each country. Finally, we talked about how we can encourage people to make a difference with AI from Bolivia to South America. At this point, I realized that we do not only need the support of the Government, also the international investment is really important and thanks to the globalization and the Internet we can apply to many international investment.

Panel

All panels were great but I have my favorite ones:

• The first one was about “How to Write a Great Research Paper (Panel: Nando de Freitas-Deep Mind, Claire Monteleoni, David Lopez-Paz and Martin Arjovsky)”. I understood that not only the research is enough , we need to create a Story for our papers: 1. Do not wait: write 2. Identify your key ideas 3. Tell ONE story 4. Nail your contributions to the mast. 5. Related work: later 6. Put your readers first 7. Listen to your readers.
• The second one was about “ Life of a ML Startup: Mario Guajardo (Google AI), Agustina Sartori (Founder & CEO at GlamST / Director AR Innovation at Ulta Beaut) , Martín Alcalá (Tryolabs Co-Founder & Director. Angel Investor.), Thiago Cardoso (CTO @hekima_br), Matthieu Jonckheere (Universidad de Buenos Aires | UBA · IC (instituto de calculo) and Department of Mathematics (FCEN)” Here the most important parts were the questions: What do you think is the biggest challenge for Startups in Latin America? their responses were: Equality and Equity, access to networks from Sout American people to EEUU, Education, Access to Venture Capitalism.

Showroom

• Sponsors had stands in the showroom where the attendees can get information about each company, many of companies were with the hat of recruiting, also this section was important to get excelent contacts. Here some pictures of the showroom:

Google in Khipu AI — Extra Events

Google had a strong participation in the conference, the stand of Google was always full, and they did the following events:

• Women in AI (hosted by Google), this event was for women where there was a panel about Women in AI. Here I knew an amazing woman: Maria Simon. She is an engineer, professor and Uruguayan politics. She was president of Antel, the public telecommunications company of Uruguay, between 2005 and 2008. She also served as Minister of Education and Culture between 2008 and 2010, and as Deputy Minister of the same portfolio between 2010 and 2012. Since March 2015, she has been Dean of the Faculty of Engineering of the University of the Republic, a position she also held between 1998 and 2005. She is a model to follow up in technology.

Women in AI Panel

Here with Maria Simon, a great woman in technology.

• “Workshop: Best Practices for AI Product Decisions”, were Mario Guajardo (Scientist in Google AI) organized a workshop focused in how to create a product with AI. From my point of view, it was kind of a little Startup Weekend where the attendees created a product in AI with their team (Product Owner, developer and a Data Scientist). If you want more information, then you can got to the following link: People+ AI Guide Book.

Mario Guajardo , Scientist in Google AI

• Last lecture was “Deep Learning to Solve Challenging Problems” by Jeff Dean who is Google Senior Fellow and SVP of Google AI and Health, overseeing Google’s research and healthcare teams. He gave us a keynote about Google AI, so we knew what are the main projects that Google AI was working on and the ambitious goals of Google AI to get the first company in the world in AI. (Healthcare, Education, Fintech, etc.).

My rate for Khipu AI is 5 starts! , this conference was really great. I like it so much because I knew south american people who are doing great things in AI. Also, the main Bolivian leads in AI came to the conference and we did a great representation as Bolivian Team.

Bolivian Team in Khipu AI

I am waiting for the next Khipu AI, and congrats to the organizers for this great Conference.

Khipu AI 2019 — Organizers

I see you in the next event!