Artificial Intelligence (AI)

 Artificial Intelligence (AI)

This is perhaps the number one most commonly misunderstood technology and also one which causes a fair amount of anxiety! I’m certainly not saying that it isn’t a cause for concern and that anyone seeking to use it shouldn’t be cautious. But it isn’t about building robots that will one day take our jobs or our planet!

The term "artificial intelligence," as it is used today in technology and business, usually refers to machine learning (ML). This simply means computer programs (or algorithms) which, rather than needing to be told explicitly what to do by a human operator, are capable of becoming better and better at a specific task as they repeat it over and over again and are exposed to more data. Eventually, they may become better than humans at these tasks. A great example of this is AlphaGo, a machine intelligence that became the first computer to beat a human champion at the game of Go. Go is a game in which there are more possible moves than there are atoms in the universe. This means it would be very difficult to program a computer to react to every possible move a human player might make. This is how conventional, programmatic games-playing computers, such as chess computers, work. But by teaching it to play Go and then try different strategies until it won, assigning higher weighting to moves and strategies that it found had a higher chance of success, it effectively “learned” to beat a human.

Until a decade or so ago, most people’s understanding of AI came from science fiction, and specifically robots as seen in TV shows and movies like 2001, The Matrix, or Star Trek. The fictional robots and smart machines in these shows were generally shown as being capable of what we call "general AI," – meaning they could have pretty much all of the facets of natural (human or animal) intelligence – powers of reasoning, learning, decision-making, and creativity – and carry out any task that they needed to do. Today’s real-world AI (or ML) is almost always what is known as “specialized” (or weak/narrow) AI – only capable of carrying out the specific jobs it has been created for. Some common examples of this are matching customers with items they might want to buy (recommendation engines), understanding human speech (natural language processing), or recognizing objects and items when they are spotted by cameras (computer vision).

Comments

Popular posts from this blog

Fossil Fuel not dead yet

3D Printing

All is flashing in 2023 for a good year