Tekoälyn määritelmästä

  • Posted on: 26 September 2018
  • By: Juho Vaiste

"Ei määrittely auta tai ratkaise asioita." Timo Honkelan kommentti tekoälyn määritelmästä Tiedekulman paneelissa.

--

Tämä avaa omaa määritelmääni tekoälystä. Olen sanonut, että käytän ihan rehellisesti ja suoraan yhteiskuntatieteisiin soveltuvaa määritelmää, ja käyn määritelmäkeskustelua rajatusti. 

By artificial intelligence, it is usually referred to an agent or a system which can operate tasks and functions to be thought earlier to be only in the reach of human intelligence.

The autonomous and adaptive nature is essential for the AI technologies. The definition of AI is on continuous change due to the AI effect. AI effect refers to a phenomenon when as soon as AI solves a problem, the solution is not considered to be AI anymore.

--

Lisäsin muutaman lauseen. 

By artificial intelligence, it is usually referred to an agent or a system which can operate tasks and functions to be thought earlier to be only in the reach of human intelligence. Ongoing debate about the nature of intelligence limit the value of philosophical analysis of AI. It is unknown if AI agents can ever reach the human-level intelligence.

The autonomous and adaptive nature is essential for the AI technologies. The definition of AI is on continuous change due to the AI effect. AI effect refers to a phenomenon when as soon as AI solves a problem, the solution is not considered to be AI anymore.

Contemporary AI applications operate in the fields of narrow AI. These applications can provide a superior performance in narrow application areas like optimization, risk analysis or image categorization. These applications are not intelligent, but more like tools based on mathematical analysis and data science. The applications require planning, implementation and supervision by human operator.