Very well done, exactly what I feared and I think if the ancestor of this (google etc) is controlled by one side of the political spectrum, the later incarnation is just a more sophisticated, subtle form of control. 'Oh the AI is a black box, they can't be biased on the fly like google'...
Hi Callmedeno, I appreciate the feedback! Google and other BigTech do mediate on what’s appropriate and what isn’t for the public. Misinformation is often suppressed to protect us. Likewise, as sophisticated AI becomes integrated in most computer platform use there will be more opportunities to dynamically nudge the behaviour of the public.
Nice experiment, it's evidence towards what a lot of people were suspecting about the technology and the biased inputs it is given. What concerns me is the prospect of a future where people advocate off-loading all important decision-making to an AI like this - an AI that is built from the arbitrary biases of humans. And because an AI like this can be so compelling with its eloquence and "knowledge" base, people with the same biases of the AI will be advocating for giving it ultimate authority over any human.
Hi Snack Entity, yes people just willingly submitting to AI decisions in blind faith is the concern! Firstly, AI could make judgements at a high level that might disadvantage individuals. In a theoretical example a local authority intending to ease traffic congestion may use AI to assign children to specific schools, or new citizens in social housing to particular regions of a city, some children may end up at worst schools or living in areas with more crime. But the most concerning is that as we have seen with all the social media platforms, what they allow users to say and not say is influenced by their owners. Their values, their biases ultimately are inherited by their platforms and this will happen with AI platforms. If it becomes too obvious, or too biased then perhaps, like with newspapers and TV channels, people will gravitate to AI platforms that confirm their own social-political beliefs.
Why are people so willing to give up using their own minds? I could care less what the machine has to say. I will continue to make my own decisions. You should too.
Hi Theodore, you make a point I completely agree with. However the point of AI behavioural nudging is that it’s not meant to be obvious, so we may not always be aware it’s happening! It could be something minor like an online grocery store nudging you to buy another pack of chocolate biscuits. Or it may be your local government encouraging you to change your behaviour so you don’t use as much electricity during peak hours. These nudges may not all be nefarious, but I would like all people to become more aware of this.
