“CRUNK” is a useful anacronym (China, Russia, North Korea) that describes well the biggest threat to democracy and ethical behavior – some will say that our largest member state of the Western world is also a threat to our security too. However, we can probably be correct in thinking that they will come to their senses soon. The same cannot be said of the CRUNK Club!
This week Helsinki has hosted the AI FORUM 2018 with a full 2-day program (see link for AI Forum 2018 videos) on the most important developments related to AI. It has been attended by leaders from the European Commission, governments, academia and leading actors from the world’s biggest companies.
At these forums most of the talk was concerned with pointing out the benefits and what needs to be done to minimize the dangers of this important sector.
One of the relevant questions that we asked about from the speakers was “Do the benefits of AI for humanity outweigh the dangers and costs of AI?”
We all know and acknowledge that AI and Machine-learning have demonstrated exceptional performance in domains as diverse as recognizing objects in images, detecting fraud, diagnosing disease, making recommendations to customers, and catching crooks.
There are thousands of applications that are useful and important for mankind – for weather forecasting, for agriculture, for genetics, for understanding eh brain and for understanding our universe.
The benefits for mankind when used well are clear and real – but the big question is:
“Do the benefits outweigh the dangers?” – some of these dangers are listed here:
- Digital facial recognition – The Chinese are using it to follow people in ways that is said to breach basic human freedoms. Our security people, and many others, in most developed countries are also collecting facial data at public places like airports, stadiums, stations, on pedestrian streets and on roads from moving cars…
- Automated weapons like bomb carrying drones use AI to pilot themselves to destroy precise locations. Governments do this and what stops criminals?
- AI is being used for cyber-attacks on major companies and banks holding our private data, to crack passwords, to attack strategic infrastructure like power plants, power transmission networks, water systems, district heating networks, and aircraft, ship and car networks that use GPS.
- Cyber-attacks can destroy mobile phone networks and security networks.
- Rather than destroying or stealing data, an attacker might make subtle, almost unnoticeable changes to the system that would undermine the way the AI algorithm works.
- CRUNK cyber hackers and Facebook have meddled or permitted meddling with our democracies in ways unimaginable 10 years ago.
- Robots can be controlled remotely, its limbs manipulated and its cameras used to spy on users.
- AI is being used to create fake videos and photos – just like counterfeiters but without any criminal sanctions!
- Mega data being held anywhere in hard disks connected to the network are increasingly under from hackers and criminals. The more sensitive the data the more disastrous the consequences…
These are all real and present dangers for which there are no robust solutions – increasing society’s digital network dependence increases vulnerability… That is not acceptable compared to the benefits…
The conclusions of many leading AI experts here in Helsinki are that those in Europe who are involved in the development of AI and Machine Learning must learn faster than the bad people. The benefits can be reaped and we can stay ahead of the game so long as we understand that time is of the essence.
At the moment the amount of money spent by China and the US is 3 times and 6 times greater respectively than the EU combined. We in Europe still do not have any major global successes in the AI field, only relatively small public and private implementations for specific sectors and our present and planned big data networks that are foundations for the future are unfortunately open to attacks by sophisticated CRUNK state hackers and ordinary criminals. This is not a good position to be in but the thing is that we have no safe alternatives to AI…