Would you trust an AI version of yourself to make decisions for you?
First of all, let’s take all the gags about ‘intelligence’ as read before we get on with the blog. Ready? Right…
A company called Sidekik has created an AI version of President Donald Trump, which they claim can tweet and interact just like him. Based on his Twitter history, the tech can presumably analyse thought and speech patterns, using that data to anticipate how the Donald would respond to any given conversation or topic.
Their vision is for everyone to have their own AI programme – a kind of cyber Mini Me – that can get to know you so well, it can even help you to make decisions when you’re stuck for an answer. So, if you don’t know what to choose for breakfast, you can turn to, well, you for help.
In order to offer this capability, Sidekik needs to be given access to your internet data, although you can choose which apps it can plunder. So, let it scour your Facebook history, for example and it could tell you to get in touch with a friend, if recent interactions show that you’ve not been in contact as often as you ordinarily would.
The creators claim it can even just chat with you about your hobbies and interests, based on the information you’ve shared or researched.
The software is run locally on mobile phones, so this doesn’t appear to be some cynical data mining exercise by big business. But the question that still presents itself is: just because we can do this, should we? Do we actually need AI versions of ourselves?
If you’re working on AI that can enhance or support our daily lives, let us know on 0800 772 0800.