Scientist Claims AI May Already Be Conscious
Topics: Science, Technology, Artificial Intelligence
Poppy Bilderbeck
Featured Image Credit: Alamy
Topics: Science, Technology, Artificial Intelligence
A top scientist has revealed that advanced AI may already be sentient.
Ilya Sutskever is making Black Mirror: Bandersnatch feel like an all too real possibility, after the chief scientist of the OpenAI research group took to Twitter to announce the possibility of today's 'large neural networks' being 'slightly conscious'.
Sutskever co-founded the organisation in 2015 alongside the company's CEO Sam Altman and SpaceX CEO Elon Musk.
Technology may have made many advances over the last few decades, but this is likely the most exciting - or terrifying - of them all, depending on how you feel about robots slowly working their way up to the level of human intelligence.
OpenAI was first founded as a non-profit organisation to try and combat the potential risks of machines becoming conscious, The Byte reports.
Advert
It states it's mission is 'to ensure that artificial general intelligence benefits all of humanity', however, suddenly, the company u-turned and has worked in a bid to try and develop and grow AI further.
Appearing on a documentary called iHuman Sutskever claimed that artificial general intelligence (AGIs) not only has the power to 'solve all the problems that we have today,' but also 'the potential to create infinitely stable dictatorships'.
Tamay Besiroglu, a student at MIT's Computer Science and Artificial Intelligence Lab, replied to Sutskever's tweet sharing a 'recent paper on compute trends in ML' to give an insight into the scientist's claim.
On the graph, the training compute of milestone machine learning systems over time can be seen reaching a point of 'maybe slight[...] conscious[ness]' towards the end of 2020 in terms of 'large scale era'. The 'deep learning era' line is still yet to reach above the point of consciousness on the graph, which dates until 2022.
However, the data was branded by another user as looking 'extremely cherrypicked' and 'gregarious,' leaving little visible evidence to back up Sutskever's claim.
In 2019, OpenAI became the subject of scandal after rumours surfaced that the organisation had created a dangerous generator which spurned out 'fake news'.
Musk resultantly left the company stating he 'didn't agree with some of what [the] OpenAI team wanted to do'.
Soon after, the organisation became 'capped profit' as opposed to previously having been non-profit.
Musk has since gone on to make serious advances in his vehicle manufacturing business, Tesla. He has also resolved that the company will do its 'best' to achieve 'decentralised control of robots' ahead of Tesla's upcoming humanoid robots, which 'might play a role in AGI'.
OpenAI's mission may be to 'ensure that artificial general intelligence benefits all of humanity,' but rather than picturing self-driving cars, all my mind is yoyoing in-between is daleks and the possibility I am simply an avatar whose every action is already pre-planned or controlled by some higher power.
If you have a story you want to tell, send it to UNILAD via [email protected]