The phrase artificial intelligence was used again in 1956 to characterize the name of a workshop of scientists at Dartmouth, an Ivy League school in the US.
During this pioneering workshop, participants discussed how computers will immediately carry out all human tasks that require intelligence, such as playing chess and other games, writing good songs and translating text from 1 language to another.
These leaders are very optimistic, even though their ambitions are unthinkable. His famous article in 1950 introduced the Turing test, a challenge to find out whether a smart machine can convince someone that it really isn’t a machine.
Research for AI in the 1950s to 1970s centered on compiling applications for computers to perform tasks that demanded human intelligence. A historical example is the application of American computer game leader Arthur Samuels to play chess.
The program was improved by analyzing the ranking of winners, and soon found playing chess much better compared to Samuels.
However, what works for chess fails to make fantastic programs for more complex games like chess and going. Other ancient AI research projects address the issue of introductory calculus, especially symbolic integration.
Years later, symbolic integration turned into a problem and the application was solved because it was no longer labeled as AI.
Compared to checkers and integration, language project translation applications and speech recognition make minor improvements.
Interest in AI has surged since the 1980s through specialist systems. Success has been reported with applications that carry out clinical investigations, analyze geological maps for nutritional supplements, and configure personal requests, such as.
Although it helps for narrowly defined problems, specialist systems are not robust or whole, and require detailed knowledge from specialists to be developed. The application does not show general intelligence.
Following the surge in AI’s initial action, research and commercial interest in AI subsided from the 1990s. And the translation application can provide a core report.
However, no one thinks that computers really understand the current language, apart from significant developments in areas such as chat-bots.
There are definite limits to what Siri and Ok Google can do, and the translation does not have subtle circumstances.
Another task of believing the struggle for AI from the 1970s is face recognition. Application at that time was not possible.
Today, on the contrary, Facebook can distinguish individuals from many tags. And the camera application recognizes faces well.
Nonetheless, this is an innovative statistical method that conflicts with helpful intelligence.
Smart But Not Smart
In task after assignment, following detailed analysis, we can create general algorithms that are applied effectively on computers, in lieu of learning computer itself.
In chess and, more recently, PC applications have conquered the winning human player. This effort is extraordinary and intelligent techniques are used, without contributing to overall intelligent abilities.
Right, the winning chess player doesn’t have to be a winning player. Maybe being a specialist in one type of problem solving is not a sign of fantastic intelligence.
The final example to think about before looking into the future is Watson, developed by IBM. Watson famously conquered the human winner on the Jeopardy TV game show.
IBM currently applies the Watson technology using the assert that it will make a proper medical diagnosis by studying all medical reports.
I am not comfortable with Watson making medical choices. I am happy to be able to shout out the evidence, but it is far from understanding health conditions and making a diagnosis.
Likewise, there are claims that computers will improve teaching by adjusting students’ mistakes for misunderstandings and known errors.
However, instructors who are broad minded are needed to understand what is happening to children and what motivates them and what is lacking at the moment.
There are many areas where human conclusions must hold, for example legitimate conclusions and launching military weapons.
Advances in computing in the past 60 years have greatly increased the work that computers can do, which is considered to involve intelligence.
However, I think we have a considerable distance before we produce a computer that can match human intelligence.
On the other hand, I’m used to autonomous cars to drive from one area to another. Let’s continue to focus on making computers simpler and more useful, and don’t worry about trying to replace us.