Prof Stuart Russell wrote the book on artificial intelligence. But that was back in 1995, when the next few decades of AI were uncertain, and, according to him, distinctly less threatening. Sitting down with Ian Sample, Russell talks about his latest book, Human Compatible, which warns of a dystopian future in which humans are outsmarted by machines. But how did we get here? And what can we do to make sure these machines benefit humankind?

In 1997, Garry Kasparov famously lost his rematch with IBM's Deep Blue, marking the first time a reigning world champion had been defeated by a program under tournament conditions. Much of the press that followed was predictably hyperbolic, with headlines questioning whether a "Terminator scenario" was just around the corner. Twenty years on, the potential danger posed by powerful AI is in the spotlight once again. It's a concern that leads to the fourth and final question of this mini-series: if we cannot align AI with our own goals and values, do these systems need an off switch? To help explore this issue of AI safety, Ian Sample calls on a trio of experts, including the University of New South Wales's professor of artificial intelligence Toby Walsh, the University of Oxford's Professor Sir Nigel Shadbolt and Dr Yasemin J. Erden from St Mary's University in Twickenham.

In October 2017, researchers at Google DeepMind published a paper on an artificial intelligence (AI) program called AlphaGo Zero. Unlike previous incarnations of AlphaGo, this updated version mastered the game of Go through self-play alone. Talking about the achievement, lead researcher David Silver explained that AlphaGo Zero had invented "its own variants which humans don't even know about or play at the moment." And it's here that a new and exciting use for AI comes to light. Could it be that AI might teach humans about the world around us?

When it comes to artificial intelligence (AI), people often obsess about creating "human-level" AI. Some think it will take decades, some centuries – and others say it may never happen at all. But is this the best marker to aim for? Or should we broaden these horizons to include the full expanse of intelligence? And ultimately: what kind of intelligence will AI researchers create?

The media is full of stories of the hopes and fears over rapidly evolving artificial intelligence (AI) technology. But it's not just pundits debating the pros and cons of what AI might one day achieve: among AI researchers themselves there is plenty of disagreement over what the future holds. Instead of trying to predict what will happen in decades to come, this mini-series will look at the here and now, and pose four questions to experts about AI in 2018. First up: what are the key challenges AI researchers are wrestling with today? To help Ian Sample try to identify some of these hurdles we hear from Prof Manuela Veloso, head of machine learning at Carnegie Mellon University in Pittsburgh, who believes that the key to success is data, data, and more data. Joining them both is Arizona State University's Prof Subbarao Kambhampati, who highlights the need for reasoning as well as recognition in machine learning systems.

In 2014, a new research and outreach organisation was born in Boston. Calling itself The Future of Life Institute, its founders included Jaan Tallinn - who helped create Skype - and a physicist from Massachusetts Institute of Technology. That physicist was Professor Max Tegmark. With a mission to help safeguard life and develop optimistic visions of the future, the Institute has focused largely on Artificial Intelligence (AI). Of particular concern is the potential for AI to leapfrog humans and achieve so-called "superintelligence" – something discussed in depth in Tegmark's latest book Life 3.0.

On Monday 20 April, a crowd gathered in Kings Place to hear a discussion on the future of Artificial Intelligence - or AI - as part of our Brainwaves Series, supported by SEAT. How do we define human intelligence? How close are we to reaching it with machines? And what happens when these machines start taking our jobs? To discuss all this and more, Ian Sample was joined on stage was Anil Seth, professor of cognitive science and computational neuroscience from the University of Sussex, Maja Pantic, professor of affective and behavioural computing at Imperial College London, Anders Sandberg, senior research fellow at Oxford University's Future of Humanity Institute, and Alan Winfield, professor of robot ethics at UWE, Bristol.