I recently reread an article from 2018 on "Deepfakes for good", which describes AI being used to generate synthetic brain MRI images. This extremely rapid scan creation allows for improved training and understanding of tumours, with a view to more efficiently approximating and beating conditions like Alzheimer's.
Nothing tops a skilled surgeon with a steady hand and an encyclopedic understanding of their craft. But all experts are constantly learning, and what better source of knowledge could there be for a brain surgeon than 100 billion extra slides of unique brain tumours to crunch through?
Now, there are a number of ways this sort of power could be abused or manipulated by greedy and malicious parties. So the right pressure and regulation is always required (see my previous post on the Google / health data fiasco).
But, on the whole, one of the most valuable - when in the right hands - uses of artificial intelligence is the simulation of medical conditions, tissue/organ slides, genomes, and realistic natural phenomena.
It's clear that we need to drastically speed up our computational potential when it comes to dealing with these types of simulation, as even a single raindrop hitting a leaf is more complex - particle by particle - than anything our best supercomputers can currently replicate with any true realism.
If the most powerful future quantum systems are carefully protected from the interests of greedy surveillance capitalists and power-hungry destroyers of privacy, there's a good chance we may actually be able to advance our healthcare by orders of magnitude and crush the most pesky terminal illnesses.
Sadly this outcome is improbable. We're more likely going to end up with various private companies sneaking their way into third party access to our most sensitive health data, abusing the progress of artificial intelligence for their own profits, and even turning our most intimidate data (blood results, heart rates, skin conditions, organ weaknesses, psychological disorders) into nodes belonging to a sociopathic marketing machine designed to toss us from one planned outcome to another like marionette dolls.
What might this puppeteering look like? I take a stab at predicting future scenarios here and here.
Is The Data Exhaust resigned to accept this as inevitability? Not yet! If we strive to explain, amplify, and deride the negative ramifications of data misuse and corporate/government psychological operations, we may yet be able to predict and prevent some of the most disagreeable aspects of the future.
Or we might accidentally give despots a few ideas they hadn't considered yet.
—
If you enjoy The Data Exhaust and the topics of surveillance capitalism, cyber security, life with technology, and information privacy, we gladly accept tips and donations, however small, in the form of Bitcoin:
3MB1z7MswieMiYYSFzMnwabPgy8JaptMgV
