As companies scramble to produce Artificial General Intelligence (AGI) – systems with human levels of intelligence – machine consciousness could soon be achieved without its creators intending it or being aware of it. The more we understand about consciousness before that point, the better chance we have of avoiding harms to humans and AIs.
That is why there is a growing global community dedicated to understanding machine consciousness, which could well hold the key to developing safe AI. The principles to guide anyone researching in this area are being established and 2025 could be a breakthrough year, adding to the recent research developments outlined in your article.
We don’t know whether machines can become sentient, but, as Anjana Ahuja points out, we can’t be confident that consciousness will never materialise in artificial brains. If it does happen, then we should be ready to detect it, measure it, and understand how it manifests – or prepare to face potentially severe consequences.
Signed:
Dr Daniel Hulme
Calum Chace
Co-Founders, Conscium
18 November 2024
