Artificial Intelligence
| Artificial intelligence researcher Marcus Hutter believes he has found an equation for artificial general intelligence. His team is now testing AIXI by playing video games. |
For artificial intelligence resarcher Marcus Hutter, others working in the field have given up on creating general intelligence, and instead have focused on related but more limited concepts. Hutter though argues that mathematically defining intelligence is not only possible, but crucial to understanding and developing super-intelligent machines.
For Hutter and his team, based at the Australian National University, the recurrent theme they found in their work on intelligence is that, "Intelligence is an agent’s ability to achieve goals or succeed in a wide range of environments."
Referring to the equation above, Hutter writes,
Imagine a robot walking around in the environment. Initially it has little or no knowledge about the world, but acquires information from the world from its sensors and constructs an approximate model of how the world works.
It does that using very powerful general theories on how to learn a model from data from arbitrarily complex situations. This theory is rooted in algorithmic information theory, where the basic idea is to search for the simplest model which describes your data.
The model is not perfect but is continuously updated. New observations allow AIXI to improve its world model, which over time gets better and better. This is the learning component.
AIXI now uses this model for approximately predicting the future and bases its decisions on these tentative forecasts. AIXI contemplates possible future behavior: “If I do this action, followed by that action, etc, this or that will (un)likely happen, which could be good or bad. And if I do this other action sequence, it may be better or worse.”
The “only” thing AIXI has to do is to take from among the contemplated future action sequences the best according to the learnt model, where “good/bad/best” refers to the goal-seeking or succeeding part of the definition: AIXI gets occasional rewards, which could come from a (human) teacher, be built in (such as high/low battery level is good/bad, finding water on Mars is good, tumbling over is bad) or from universal goals such as seeking new knowledge.
| Related articles |
. AIXI is incomputable, and has to be approximated in practice. In recent years, Hutter and his team have developed various approximations, ranging from provably optimal to practically feasible algorithms
Currently they are at a toy stage: using the approximation to teach AIXI to learn to play Pac-Man, TicTacToe, Kuhn Poker and some other games.
According to Hutter,"The point is not that AIXI is able to play these games (they are not hard) – the remarkable fact is that a single agent can learn autonomously this wide variety of environments."
For the games, the system is given no prior knowledge of the games, or even told the rules. It starts as a blank canvas, and just by interacting with these environments, it figures out what is going on and learns how to behave well. This is the really impressive feature of AIXI and its main difference to most other projects.
Unkike Watson, which was originally built just to win at Jeopardy!, or Deep Blue for chess, AIXI is able to be more general and flexible. While still in early development, AIXI and other work like it show that artificial intelligence has the potential to be a much more general system.
SOURCE Marcus Hutter via The Conversation
| By 33rd Square | Subscribe to 33rd Square |
0 comments:
Post a Comment