Artificial Intelligence is the go to for everything these days. As time goes by the use of AI is getting more and more not only limiting to offices but also people trying to use it in their day to day lives. For e.g. here in India we can use Amazon Alexa + Amazon Fresh. Alexa is the AI assistant, where you can give your commands and it will take action accordingly to order your list from Amazon Fresh which is the shopping portal. If it is an item which needs to be purchased at regular intervals, Alexa takes the command and does the needful.
With these tools you can actually create your AI powered monthly grocery list based on what are the regular monthly items bought. These days when I travel I take AI help to plan my trip, to a great extent it is accurate but sometimes it gives misinformation also. During my trip to Japan, on couple of instances I did get wrong information. My conclusion is that even though it is giving a lot of ready information, it cannot be taken as 100% accurate, and I would not want to rely on it for things which are of importance and I would better keep it more human.
Today we are still using AI for small things, mostly queries but in future as this develops more and more, may be we will use them for more complex actions, which to an extent are happening now also. So, when AI goes wrong, who is liable? Is it the user, the developer or the company deploying it. As of now if there is anything that goes wrong with a product or a service, we know who to exactly catch hold of. It can be the manufacturer or the seller, but if I have an AI working for me then how can I make the AI responsible for the mistake and in court of law who will come and stand.
I think in these cases, it will be important to retain the commands given to the AI to prove it in some way that the system defaulted. We have always been asking what can AI do, but soon we will also have to ask what if AI goes wrong, who will be accountable. The tool may be developed by someone sitting in one corner of the world and you may not be able to touch that person in any way, so then what? The good news for Indians is that the Data Security Council of India is working on a framework to enforce a liability for AI going wrong. After all AI is just not a tool anymore, it is working like an agent, so someone has to take accountability of wrong actions taken by it.
Thank you for visiting my blog. 馃懠馃徎馃懠馃徎馃挅馃挅馃尮馃尯馃尭
