This is the second blogpost in a series on Artificial Intelligence and Human Rights, co-authored by: Christiaan van Veen (Center for Human Rights and Global Justice at NYU Law) & Corinne Cath (Oxford Internet Institute and Alan Turing Institute). Why are human rights relevant to the debate on Artificial Intelligence (AI)? That question was at the heart of a workshop at Data & Society on April 26 and 27 about'AI and Human Rights,' organized by Dr. Mark Latonero. The timely workshop brought together participants from key tech companies, civil society organizations, academia, government, and international organizations at a time when human rights have been peripheral in discussions on the societal impacts of AI systems. Many of those who are active in the field of AI may have doubts about the'added value' of the human rights framework to their work or are uncertain how addressing the human rights implications of AI is any different from work already being done on'AI and ethics'.
Researchers presented the fish with two images of human faces and trained them to spit at one of them (archerfish are known for spitting water to catch flying prey). Once the fish learned to recognize that face, researchers presented them with the learned face and dozens of new faces. The fish reached an average peak success rate of 81 percent in this experiment.
The number of online human rights abuse cases in Japan in 2016 grew 10.0 percent from the previous year to 1,909, hitting a record high for the fourth straight year, the Justice Ministry said Friday. The overall number of human rights violation cases for which actions were taken last year came to 19,443, down 7.4 percent. Of the online abuses, privacy violations such as the disclosure of personal information totaled 1,189 cases. There were 501 cases of defamation. A total of 1,789 cases of the online human rights abuses were resolved, including 326 cases in which the deletion of abusive language and information was requested.
My colleague Ahmed posted yesterday an article that had appeared about a new record for ASR performance. Cortona had reached an ultra-low 6.3% error rate. Human voice dictation error is in the 4–6% level, which means we're just on the cusp. In May, I made a prediction at SpeechTek that within the next twelve months, we'd surpass human error rates in ASR. Of course, the caveat is that it would be a well trained and tuned ASR and for a specific individual.
I believe the three ways in which AI can enhance human innovation -- namely, creating space for innovation, generating novel patterns and democratizing creativity -- have the potential for real impact on the business world. However, a few qualifying words are due before we end the discussion. True human creativity and innovation are only possible when there is a solid basis to build on. Moreover, the more robust the substrate on which you build, the more ingenious the innovation will be. The more educated and trained a person is, the more they can create.