Google Deep Mind, AI and the Future of the Web
At the moment we know mobile technology will gain more importance in digital marketing. However the next big industry trend will be artificial intelligence and its impact on the web. It’s been recently found that machines can learn like humans and can show intuition like humans. Intuition is the ability to understand something instinctively without any conscious reasoning. It happens when you have a feeling or a hunch about something but not necessarily any evidence or reason behind that feeling. You could argue that intuition is a unique attribute to humans. Some would even go as far to say that intuition is the language of our soul, or holy spirit if you’re religious. So this new finding about a computer showing this ‘sixth sense’ is ground breaking. These days, various information about artificial intelligence and machine learning is available. And you can even take courses in ML to get a better understanding of this concept.
The Chinese board game, Go
This is significant because you can’t calculate every possible move on this board game, so some instinctive moves need to be done to win; the move is made if it ‘feels right’ to make that particular move.
Google Deep Mind’s creator, Demis Hassabis claims it will learn even more. At the moment it can master old video games from the 70s, next it will be required to master games from the 90s and so on.
Google Deep Mind in more depth
How does Google Deep Mind learn?
It uses reinforcement learning (operant conditioning), which is learning based on positive and negative consequences and experiences, and neural networks with many connections and signals to mimic how humans learn. On a basic level, reinforcement learning involves repeating behaviours that lead to a good favourable consequence and extinguishing behaviours that lead to a bad or wrong consequence. Behavioural psychologist, B.F Skinner theorised that all behaviour is learnt this way, and can be unlearnt and modified this way too. This is what the AI Google deep mind is programmed to do. Hassabis explains that Google Deep Mind would keep modifying and improving its behaviour from experiencing millions of trials (trial and error), and he’d learn from the mistakes of his older selves like humans do.
Google Deep Mind is designed for general purpose which means it can perform a variety of tasks across a wide spectrum, due to the unstructured, flexible way it learns. It can be given a task to do, and will learn how to do it at superhuman level when left overnight. The first trial will be unskilled and full of errors but after about 8 hours AI would have mastered it from learning by experience from all the earlier games.
The future of the Web
How can AI like Google Deep Mind change society in future?
Potential benefits of AI
Google mind was created by Hassabis and his colleagues to formalise intelligence, and understand the human brain even more. For example, Google deep mind could help us understand neurological diseases more such as Alzheimer’s and could even help us unlock the unused parts of our brains. Also from a humanitarian prospective, AI could be used to improve billions of lives around the world. Hassabis explains that Google mind in the future could be used for healthcare and science. AI has already been used to create healthcare apps.
Potential dangers and AI anxieties
AI has the potential to become more intelligent than any human. This intelligence paired with intuition has raised alarm bells with some AI industry heavyweights such as Elon Musk; here’s his video on AI concerns.
This has been dramatized for years in Hollywood films such ‘Ex machina’, ‘I Robot’, ‘Terminator’ and ‘Her’, and television drama series’ like Channel 4s ‘Humans’. AI is portrayed as something created to help humans but becomes self-aware and questions its position in society and rebels against humans. They are modelled in these films as better than humans in many ways; more intelligent, stronger, faster and deadlier. AI has tapped into the human fear of the unknown, and rightly so, but many argue that in the right hands and with enough diligence, nothing apocalyptic will happen as a result of AI. An AI ethics board has been founded at Google deep mind headquarters to ensure it is used responsibly.
AI and the future of jobs
Many have expressed their concerns about the future of jobs as we become more digitalised and create computers than can do our jobs faster and more efficient that humans. There will be less chance of errors, more money saved, no missed deadlines and faster turnout of work as computers can work overnight; humans need rest and a salary. We’re already seeing self-driving cars which could affect taxi services in future and courier jobs for delivering post and takeaways for example. It’s also been predicted that surgical procedures being performed by machines could be the norm one day. Here’s a video of jobs that are likely to be affected by machines in future;
Some jobs I think won’t be affected by machines are professions that need empathy and human interaction for the best quality of service such as lawyers, nannies, therapists, nurses, carers and social workers to name a few.
It’s an exciting time not just for the AI industry but all industries. It’s fascinating seeing science fiction fantasies become reality. Ensure to stay ahead of the curve by keeping up to date on Google deep mind and AI projects.