Tech Tent: Autonomous cars and AI doctors

Rory Cellan-Jones
By Rory Cellan-Jones
Technology correspondent
@BBCRoryCJ

Was this the week that the space age vision of a car that drives itself became a reality? And are claims that artificial intelligence can transform healthcare a bit overhyped? On this week's Tech Tent podcast we explore the potential and limits of technology in health and transport.

Driving into the future?

This week we woke up to the fact that autonomous cars could be with us sooner than we thought.

That was the message from John Krafcik, chief executive of Waymo, the self-driving car division of Google - or Alphabet as we must learn to call it. At the Web Summit in Lisbon he announced that fully self-driving Waymo cars with nobody behind the wheel were now in operation in Phoenix in Arizona.

Now, there will be a Waymo engineer in the back seat - and the cars will only operate on designated routes. But this still looks like a big step on the road to the autonomous driving future.

It is not just Silicon Valley making big claims about the progress of this technology.

Reuters

Waymo's autonomous cars are on the roads

Former General Motors executive Bob Lutz made a stir this week by predicting that the era of the human-driven automobile would be over in 20 years. He says there will still be a few sports cars and off-road vehicles, but for most of us driving will be a forgotten skill.

Our special guest Prof Paul Newman, founder of the autonomous vehicle firm Oxbotica, naturally shares that vision of the future.

But we also hear a dissenting voice - the transport writer Christian Wolmar thinks such predictions of the advance of autonomy are laughable.

He says they fail to take into account all sorts of legal and regulatory issues to which there are no easy answers, and he points to one example: "If you stand in front of a driverless car it cannot drive at you, it cannot kill you so it has to stop."

He foresees pranksters causing gridlock by standing in front of driverless cars, or criminals stopping them to rob the occupants.

Can tech transform healthcare?

Technology has always promised to transform healthcare, but often it has failed to make quite such a difference as its promoters claim. On the podcast, we explore whether the smartphone revolution and advances in artificial intelligence will mean things are different this time around.

One entrepreneur who thinks we are now at a tipping point for health tech is Elina Berglund. She has had a remarkable career change, from physicist in the Nobel-prize winning team that discovered the Higgs Boson particle to creating a birth control app.

This week she received $30m (£23m) of funding for Natural Cycles, which she says combines centuries-old mathematics with more recent medical research about the likelihood of conceiving at particular times and puts it all into a state of the art smartphone app.

She tells us that advances in technology are going to mean a big change in the way we see healthcare: "I do believe we will go from the current system of treating already existing diseases to understanding our bodies at an individual level, and then you can prevent health issues in the first place - and that is really the future."

We also meet a doctor and a scientist who have spent six years on an artificial intelligence project that is now improving the care of patients with brain tumours.

Antonio Criminisi, of Microsoft Research in Cambridge, and Dr Raj Jena, from the city's Addenbrooke's Hospital, have been getting a computer to understand what a brain tumour looks like, giving it some of the doctor's knowhow.

While there is much talk of computers becoming more efficient than humans at analysing scans, Dr Jena insist this AI project is about augmenting his skills rather than replacing him. "It's certainly not going to make me redundant, that's for sure."

He says that preparing a brain tumour patient for treatment, which involves laborious examination of a series of scans, can take him well over an hour but the machine can do that in three or four minutes. "What's really exciting is that it allows me to work faster and to get patients on to treatment more quickly."

But Julian Huppert, the scientist and former MP for Cambridge, warns that when it comes to AI and health "we really are still in the hype cycle".

He now runs a university forum that looks at how artificial intelligence could transform healthcare, and says the speed of progress will be limited by the problems the National Health Service faces in sharing data.

"We know how much the NHS has failed to transform its IT system. Billions of pounds have been put in and we still don't have a proper secure and accessible system for health data. If you don't have that you're really limited as to what you can do with an AI."

Then there is the question of who will own the advances in medical science generated by AI. Will Google's DeepMind or IBM's Watson win Nobel Prizes for medicine and generate huge profits for their shareholders? Will patients who hand over their data get any reward for doing so?

There is no doubt that artificial intelligence could offer extraordinary benefits, from accelerating the development of new drugs to the earlier detection of diseases. But just as with the technology behind autonomous cars, there are plenty of old-fashioned ethical, legal and practical problems that will have to be solved first.