[TGE] 6 week. Results of the conference "Opentalks.AI": new business partners. 5 February – 11 February
Last week, the "OpenTalks.AI" conference was held under the auspices of the GraphGrail Ai team . The leader and ideologist of the company, Victor Nosko, received many new proposals for cooperation at this event.
The largest conference on Artificial Intelligence "Opentalks.AI" was held in Moscow in the office towers of the Mail.ru Group. The event was held over the course of 3 days with each day devoted to science, business, and ethics.
The conference was attended by top speakers, both from Russia and from abroad (representatives of Nvidia, Samsung).
The first day belonged to the representatives of academic science who are engaged in machine learning and Artificial Intelligence development. In particular, Constantine Vorontsov, the author of books on Artificial Intelligence and various methods of machine learning, (previously worked in Yandex). Vorontsov now works in his own company dealing with AI in the field of text analysis. In his report, Constantine spoke about the fact that many problems can be solved without resorting to the use of complex algorithms of machine learning. Upon discussion of the speech, Victor Nosko posed the following question - "You say that you can make certain dictionaries for a majority of tasks based on keywords, but how can we use such dictionaries when we need to define irony or sarcasm?". Constantine agreed that these are complex and unsolved problems and answered that work is being carried out in this direction.
Burtsev Michael (iPavlov.ai), a representative of a group of developers from the MIPT dealing with the framework for building chat bots also gave an interesting speech. In particular, an extremely complex scheme was developed for handling calls to calls center based on rules. The downside of this approach is the cost of maintaining its efficiency as it is constantly growing, becoming more complex and programmers need more time and money to maintain such a complex system at the same quality level. The GraphGrail Ai project can, with the help of certain iPavlov.ai solutions (they are in the public domain), create a kind of convergence to provide data and to conduct integration in the GraphGrail Ai platform.
Burtsev about iPavlov.ai
Another speech was given by Sergey Shumsky, an expert in the field of machine training and artificial intelligence, vice-president of the Russian Association of Neuroinformatics, leading expert in direction of NeuroAssistanty at NeuroNet.
Particularly worthy of note is the presentation by Andrey Tarasov, head of the Intelligent Data Analysis Laboratory at Sputnik (our domestic search engine). The team analyzes tone and aspect-Sentiment analysis on various topics (drinks, food, service, etc.) trying to understand what is important to people. Many problems that the team faced during the course of their work can be solved by the GraphGrail Ai platform (similar problems are described in detail in our Whitepaper). After the speeches, Andrey Tarasov and Victor Nosko discussed partnership between GraphGrail Ai and Sputnik in terms of PR and applied problems in text analysis.
During a coffee break at the conference, Victor talked with Vorontsov, who did not deny a joint photo. :)
Photo with Vorontsov - author of courses and lectures on AI.
Constantine Vorontsov is a PhD in Physics and Mathematics, Senior Researcher of the Computing Center of the Russian Academy of Sciences, Professor of the Department of Intelligent Systems, Moscow Institute of Physics and Technology, Associate Professor of the Department of Mathematical Methods of Forecasting of the Moscow State University. One of the creators and administrators of MachineLearning.ru
Results of 1 day of the conference. Victor held 5 talks, including with the speakers of the conference, representatives of projects, and agreed on a possible partnership, both in terms of PR and cooperation in the field of AI. In particular, GraphGrail Ai and Skychain have agreed on cooperation in business development.
Points of cooperation include both analysis of text data, and attracting customers to platforms.
GraphGrail Ai is a decentralized AI text analysis platform.
Skychain is a platform for analyzing medical data on blockchain.
Vorontsov talks about the tasks of natural language processing for businesses
The second day of the conference was dedicated to business, including the use of Artificial Intelligence in medicine, retail, finance, industry, and the use of machine learning in cybersecurity. Afterwards, there was an investment session. David Yan (ABBYY) opened the second day of the conference. David told about his company and about the future of Artificial Intelligence.
David Yan on the company ABBYY
ABBYY is present in many countries and has various solutions for businesses, including Artificial Intelligence (smart column and fact-analysis for business). Victor spoke with David Yan, offering him partnership to become an adviser to GraphGrailAi.
Samsung stand
The presence of the Lawyer.Ru startup was also of great interest for GraphGrailAi. Lawyer.RU is a legal service using a trained system that responds to legal questions. Previously it was believed that such a system could not be created, as all legal issues are extremely diverse and have many nuances. However, the solution was achieved through initial clustering, and, then, by training the neural network to answers simple questions.
There was also an interesting report from the Rostov project UkitAi during the startup session. The service allows redesigning any site using artificial intelligence that is already trained on the best examples of aesthetic sites with good conversion rates. One of the developers of the system is our colleague from Mehmat SFU - Roman Steinberg.
UkitAi service
Ethical issues were discussed at the Medical session of the conference, including issues, such as who is responsible in case Artificial Intelligence advises certain drugs that can harm the patient. In fact, all doctors now use Artificial Intelligence, using searches in Google or Yandex, but it is necessary to have a more convenient tool, such as a decision-making assistant when setting a diagnosis or giving personalized recommendations on drugs for patients (to avoid drug incompatibility and side effects), and also the problem of the analysis of medical records. After the speeches, Victor talked to one of the speakers of the medical session (a representative of the medical company Element Lab), offering cooperation in terms of creating a database of assistance based on meanings in texts describing drugs and their dangerous side effects.
The main problem of the AI market is the need for a lot of marked data.
In the investment session, a partner of the Flint Capital Foundation gave a speech and revealed that in 2017, the volume of investments into Artificial Intelligence had dramatically increased, and he told about the problems of business adaptation to AI, based on research of the Forrester Research company and the Gartner analytical company. Flint Capital concluded that the main barriers to the adaptation of AI in 2016 are the lack of data, the lack of data-science specialists, and the high cost of developing solutions. Another significant trend in 2017 is an increase in the number of public data sets, in particular, services such as Kaggle Datasets, UCI Machine Learning Repository, KDnuggets Datasets. However, data sets need substantially more for specific industries and for solving specific business tasks.
The report also covered the statistics of investment growth in AI companies. In 2017, these investments almost doubled. First of all, corporations, business angels, grants, etc., all invested in AI.
A very important part was missing in the report of the Flint Capital partner, namely the ICO factor in the statistics on investments (in 2017, for the first time, investments attracted in ICOs exceeded venture capital). A discussion arose about the competition between investment funds and ICOs, in the process of which it became clear that funds are now still working in the old paradigm and are not invested in ICOs.
After the end of the 2nd day of the conference, Victor also met with representatives of three Russian venture capital funds (Runa Capital, Ruvento and NeuroNet) and representatives of the foreign company Mangrove Capital Partners, David Varokie, who are still negotiating participation in the GraphGrail Ai TGE.
The third day was dedicated to the ethics of using Artificial Intelligence. Representatives of academic science, philosophers from the Russian Academy of Sciences all spoke on the issue. The issues discussed concerned the responsibility of advice that AI could, for example give in family matters. Will this responsibility fall on the developer? Is there a subject and object of responsibility? How to teach AI ethics, understanding what is ethical for a person, and what is not? In fact, there is such a problem, the solution to which is present in rudimentary forms in the so-called "Moral Machine" (https://golos.io/ru--bespilotnik/@bitcoinking/mashina-morali-reshite-kto-dolzhen-zhit-igrovoi-simulyator-mit).
The visit of the Opentalks.AI conference by Victor Nosko was very successful for GraphGrail AI. The company received many proposals for cooperation and new partners have emerged. In addition to the above-mentioned business agreements, GraphGrail AI is conducting talks with AI Forge, a company that develops computer games in the field of complex and non-linear game dialogues. Other talks are also being held with Samsung on the presentation of the GraphGrail AI platform for cooperation in the new Ai center in Moscow
See you in the next newsletter!