Source: Deep Learning on Medium
What Should We Expect from AI and What we Should not!
In my last medium story I talked about AI and AGI (Hey if you didn’t read it here’s link to read it https://medium.com/@sirajmunir93/a-in-depth-introduction-to-ai-and-its-classes-d1b85747307e?). Now today’s agenda is to clarify some issue that are related to AI, specifically growth of AI.
Things were started in about 90’s when GPS (General Problem Solver) was created (1959) although at that time it was a big achievement as GPS can solve problems like Tower of Hanoi. What it does basically it uses horn clause then from clause it makes directed graph and solves the problem. If you don’t know about Horn clause don’t worry I’ll add suitable link so you can get an idea of it. In the same era fifth generation computer systems were introduced. AI was making progress and then people started expecting more from AI. People thought we will surpass our problems with AI. But there expectations failed badly. I’m not criticizing the power of AI, I just want to show what should we expect from AI and what we should not. Over expectation always cease towards bad end or disappointment. The picture that I included in start of my story shows AI winter. AI winter? it’s human nature when a person doesn’t get what he/ she is expecting they got disappointed and tries to wind up the mess. Same stuff happened with AI in previous decade when AI failed/ was not enough mature to fulfil the expectations. People started to think of AI as bad business. We humans are greedy literally we are: Whenever we invest our time/ money/ efforts we want positive outcome from it but it doesn’t workout in real world. Sometimes you pass sometimes you fail this is bitter reality to be realized well.
Alright, coming back to core topic we faced our first AI winter was started in 1973. But researchers didn’t lose their hope and worked hard until Expert Systems were developed/ introduced. Once again hopes awaked and vision of Utopia was near to achieve. Then in 1990 again AI community was hit hard by one more winter. Reason: “expert systems were good at specific tasks and their hardware maintenance was a big burden on company’s head”. This cause slow down in deployment cycle thus we faced second AI winter. The following picture well explained AI winter.
Woop! well it’s AI summer now!
Rewind what we have achieved so far after second winter. Here’s tentative list of areas maybe I forget adding some areas. So I apologize in advance!
- Intelligent Agent Systems. Also known as Multi Agent Systems.
- Deep Blue.
- Predictive Systems.
- Knowledge Based systems.
- Semantic Web.
- Evolution of Machine Learning and Deep Learning.
Concluding story: Lesson we learned from history don’t over expect from AI. We have got great success in task specific AI/ AI model trained for specific task (Classification, Regression etc).Current level of AI is not a generalize enough and for technological Utopia we need generalized version so called AGI (Artificial General Intelligence) or more than it. Again AGI doesn’t guarantee us Utopia or Dystopia as we are on our towards it. I hope this story may help you out to understand what we should expect from AI and what should not.
That’s it for now! I hope you like my story (don’t hesitate to give a applause).
Get connected https://www.linkedin.com/in/siraj-munir-2b59a7b5/
Ping me Siraj Munir
#AI #AGI #MyThoughts #MAS #KnowledgeBase #ML #DL