Following on from my resolution at this year’s IATEFL conference to attend conference talks on themes that I would normally avoid, I decided to attend Wilim’s talk on AI - something I know little about beyond the names of a few commonly-used tools.
Wilim started off with a balanced statement about the dangers and “wonders” of AI and the lack of preparedness in people’s minds for the impact this will have.
Wilim posed the question of what AI is and offered his definition linked to processes and systems and introduced three terms that were new to me:
1. Narrow AI – AI able to do one job
2. AGI (artificial general intelligence) - AI doing more jobs and understanding context)
3. Superintelligence - AI exceeds human intelligence in a way only imaginable in science fiction.
Wilim asked us to get past sensationalism linked to AI and said there will always be gaps in AI and that, when we receive an answers from AI, we must be cautious of any bias we have in the answers we received and the extent to which we immediately take these answers as accurate.
AI generates responses based on patterns but fails to understand content or context. Wilim provides the example of Chat GPT as generative AI.
This was a hugely interesting talk in which I learned a great deal about a subject I have traditionally avoided. It’s up to us now to think about how to use this.
Comments