A few days ago, I had a conversation with my 9-year old niece, where she seriously doubted the notion that computers could be smarter than humans. She argued that machines are only capable of solving specific problems assigned to them and can't be creative. I tried to explain to her about AI without undermining the credibility of her source of information ( an elementary school teacher).
I believe it has devastating consequences when we fail to provide humans with accurate information about AI from an early age. I would like to hear from experts in different fields, which is the best possible way to address this issue.
How do we explain to children that:
- The first time in human history, we aren't the smartest creatures on the planet Earth.
- AI is neither good nor bad, and it will fundamentally transform our lives in the near future.
We don't usually teach children about AI. The closest most will get to an AI is a Roomba, one of their programmable toys, or targeted ads on the internet. I also do not believe that a computer can be smarter than a person as we program them to be able to solve certain problems.
I think it could be easy to use examples such as areas of application of AI like Smart home, smart city, industrial, etc. by stressing that AI helps humans in making life easy, i.e., instead of employing a watchman you can install alarm & CCTV to guard your home. AI is an algorithm written by a human being, so it cannot be smarter than human beings. AI is a set of instructions performing what it has been instructed by a human being to do!
AI these days primarily is Artificial Narrow Intelligence. According to Andrew Ng, Artificial General Intelligence is very far and may take century to be truly realized. According to Jack Ma, AI can't have wisdom. So, your statement that AI can be more smarter than human is quite subjective and it depends upon the problem in hand. However, the objective of AI is to facilitate humans. There are a number of potential applications of AI that you can search over the internet and convince your kid about its importance.
I do agree that there are case studies about AI misuse such as Deep Fake, biaseness issues. For more details see Andrew Ng course "AI for every one" on coursera.
That we are still confusing about this new technology and understanding its unlimited boundaries and these children will take it to a next level in the soon future
First of all, I want to say that AI is good.
Electronic music didn't make people stop playing the piano. Streaming of video didn't stop people from going to theaters (Although theater equipment has greatly improved). Email didn't replace the thoughtfulness of a handwritten letter. Technology is often additive (not just disruptive).
Similarly, AI will enable a new, easy and faster way of doing things, they are tools or machines (though mostly software).
To me, they are not smarter than human begins (their creators). They are just able to get things done faster, effectively and efficiently.
Just try to compare the era of farming with simple farm tools (cutlass and hoe) to when mechanized farming (tractors, planter, and harvesters) started due to the rapid increase in population.
Human needs are increasing daily, conventional means of solving them can't meetup. AI is the machinery of the 21st century, which will be used to solve human needs.
When one talks about AI they are often talking about a specific machine learning algorithm, Deep Neural Nets in general. These algorithms or processes are not smart or creative. The whole conversation is hard to have without a firm understanding of the AI algorithms or at least partial derivative analysis and calculus. The basic idea that any mathematical function is intelligent is misleading. Identifying cats in a photo doesn't make an algorithm(which is just a series of mathmatical step executed in sequence) smart, identifying which combination of drugs/diet could cure a disease would also not make a computer intelligent. Seeing that the first example is really just looking for patterns in arrays of data and the second an optimization function. These algorithms mimic intelligence with trial and error, hence the word artificial. The premise of this question should be reviewed. A good way to think about AI is to visulize an optimization function where you have a few input variables and you want to optimize your output. Excel can do such a function, this is basicaly a simple way to think of AI algorithms. Most AI algorithms just do something similar over and over again, sometimes utlizing unique data transformations between steps and weight adjustments. This iterative application of basic functions mimics intelligence.
hi Mariam Berianidze ,
i would like to start with a very basic example!!!!
when a rich person called "master", hire some servants to perform certain duties. this does not mean that the master is not able to do that tasks like washing car, cleaning the house etc etc. but there can be possibility of unavailability due to some other busy schedules. so what master do? he assigns his duties to his servants so the system should get going.
Similarly, humans have the power of thinking and taking decisions. But there is a limitation of focus. Humans can not focus on a lots of activities at a time so they made machines to perform their duties. they made machines to work and act according to different scenarios. To perform certain actions, these machines are trained and the training phase is made possible with a set of instruction named as algorithms.
So, the smartness of any machine can be analysed by the set of training instructions or "algorithm" . Without that algorithm, a machine is nothing. there are also some limitations of machines like they will not act or behave against the situation by which they are not trained.
Concluding the whole discussion: "training our machine to act according to certain scenarios is known as machine learning" and this is done with the help of algorithms. these machines are task specified so they perform operations(mostly calculation based) more quickly than humans.
i hope you may find this helpful.
thanks
regards:
Muhammad Saad Amin
To me, it is important we teach children the power of data. In an AI world, the more the data, the more the knowledge discovered and this knowledge is what drives AI technology. Systems make decision from what the know. This decision could be either to the left or to the right.
We must be ready to teach students what knowledge type should be mined from a pool of data. AI for children should be first focused of knowledge representation and data mining rather than presenting them with all the statistics and computational functions that can make this world a hell for humans.
Children must understand the power of data and data mining. And that as humans, the more information we are exposed to, the more guided our decisions.
We should teach them how information is mined from data , the processes, constraints and work-ability of an AI world.
These are what children should learn in schools and home about AI.
Dear Saad Amin -I appreciate the briefing in a simpler way. I somehow agree with your pointing out that just like electricity -AI can be good servant but a bad master!!
To
Mariam Berianidze many thanks for bringing in and sharing this discussion.
Regards
Nowadays Robotics include the child kids .
REF : kickstarter.com ( Photo-Source ) .
We don't usually teach children about AI. The closest most will get to an AI is a Roomba, one of their programmable toys, or targeted ads on the internet. I also do not believe that a computer can be smarter than a person as we program them to be able to solve certain problems.
- Badges
- Science topic