Whenever there’s new technology, there’s speculation on how it will impact society, including the jobs we do. So, obviously there’s no shortage of speculation on the impact of artificial intelligence (AI).
On January 23, Girl Geeks Toronto hosted a panel on Ethical AI that offered some food for thought from those who develop AI.
In partnership with the new Toronto chapter of Humans For AI, the panel included experts in fintech, health sciences and the much talked about Uber driverless car:
- Hessie Jones | CMO of Humans for AI and Co-Founder, Salsa AI
- Inmar Givoni | Autonomy Engineering Manager, Uber Advanced Technologies Group
- Karen Bennet | VP Engineering, Cerebri AI
- Anna Goldenberg | Scientist at the Genetics & Genome Biology Lab, SickKids Research Institute
To understand AI’s impact – including how it may impact communications, it’s important to understand AI better. Here are 3 thoughts from the January 23 discussion:
- AI is not “magical.” Some people claim it is, but it isn’t. It’s machine learning, meaning the machine can only be as smart as it’s made, whether through supervised (most AI today) or unsupervised learning. That’s why ag machinery companies like John Deere have had driverless vehicles for almost 20 years, whereas a driverless car on our city streets and highways is much more difficult. The variables in a closed environment of a farm field are much less than driving down a city street where there are pedestrians and cars coming from any possible direction. In both cases engineers must account for all the variables to have a viable AI use case.
- AI does not solve ethical problems. An algorithm is just that – an algorithm. Because engineers establish the parameters in which it functions, if there is an unfortunate choice, say for example, in a car collision where either a pregnant woman would get hit on one side or a senior would get hit on the other, the technology will not make that choice between the two unless it has been programmed to do so. The discussion around this type of ethical decision is not up to engineers however to program it within the technology, but rather society or perhaps the individual user making the choice through an agreement before getting into a driverless vehicle.
- AI with intelligence makes it much more powerful. While AI gives organizations the ability to process more information than humans can (e.g. the human body alone has much more data than any human brain can process), by adding a human to comprehend the trends in the data, we can make the data work for us in new ways. An example is how organizations (like credit scoring companies) are already using data from your cell phone to see where you’re going to figure out what you may be doing, even when not purchasing anything.
So, will AI impact our jobs as communicators? Definitely.
But, just like how we couldn’t appreciate how social media would impact our jobs when it was first invented, personally I don’t think anyone knows yet (whether positively or negatively) how AI will impact communications. It’s part of what we all have to figure out.