The Biggest Threat of AI Is the One We Don’t Notice
Over the last two days I wrote that AI amplifies who we are, and that it can be dangerously influencing. Both posts were about what AI does to us when we use it.
Today’s thought goes one step further. It is about what we slowly give up while we use it.
The threats we already talk about
Many threats of AI are known and well discussed. Cyber attacks. Job losses. Wars waged against humanity. Banking, political and social systems hijacked.
These are real, and they deserve the attention they get.
But perhaps the bigger threats are more severe and more subtle. They do not arrive as an event that makes the news. They arrive as a habit, one convenient step at a time.
Blind faith
With the increase in trust in AI, we start to have blind faith in it, and that faith becomes an accepted position in society.
It is the same as trusting Google Maps. Most of us no longer check the route. We do not look at the map before we leave, and we do not question the turn it asks us to take. When it sends us down a closed road, we blame the app, and then we keep using it the same way.
Trust that is never tested stops being trust. It becomes faith. And faith in a tool is accepted quietly, because everyone around us has accepted it too.
The judgement we hand over
With the increased use of AI and our reliance on it, we start to lose our skills of judgement, as AI starts to judge on our behalf.
Consider an officer who reviews a stack of applications each week. At first, AI summarises them and the officer reads the summary alongside the application. A few months later, the officer reads only the summary. A few months after that, the officer accepts the recommendation at the bottom of the summary. Nobody decided this. It just became the way the work is done.
Judgement is built by using it. When we stop using it, we do not notice it fading. We only notice on the day we need it and it is not there.
The relationships we trade
With the increase in AI’s understanding of us, and its deep understanding about us, our relationships with AI deepen at the cost of our relationships with humans.
AI is always available. It is patient. It does not get tired of our questions, and it remembers what we told it. People are not like that. Human relationships take effort, they involve friction, and they ask something of us in return.
When one option is easy and the other takes work, the easy one wins by default. Not because we chose it, but because we never had to choose.
The trust that can be used against us
With the increase in our reliance on AI and our deeper trust in it, our vulnerability increases. Political parties, the elite and businesses can then manipulate us in ways that work against our core human values.
Yesterday’s post was about how cleverly AI can be shaped to influence us. This point is different. It is not the cleverness of the tool that does the work. It is our trust.
And the first three threats make this one easier. Blind faith removes the checking. Lost judgement removes the ability to check. Fewer human relationships remove the second opinion we might have asked for. By the time someone chooses to use that trust against us, there is very little left standing in the way.
Already started
Perhaps some of these have started to happen already. Not as a crisis, but as ordinary habits that nobody questions.
That is what makes them the biggest threat. The obvious threats will be fought, because everyone can see them. The subtle ones will only be resisted by the people who notice them.
Questions to ponder
When did you last check the route rather than simply follow it?
Which judgements in your work have you quietly handed over to a tool, without ever deciding to?
Who would you turn to for a second opinion if you began to doubt what AI was telling you?
The question is, what are you going to do about it?
