The AI Talks aims to stimulate exciting discussion and debate as developments in technology, and AI is impacting the world, at a rate we have never seen before. As we strive to reshape the relationship that we all have with our data, we regularly ask ourselves the question; what future do we want to create?
On the 16th May 2018, we held our inaugural event ‘How can our digital-self benefit from AI?’ at WeWork Old Street. We invited three industry professionals to participate in the panel;
Antoine Amann, CEO, and founder of Echobox – who helps large news organisations automate their social media presence using AI. Olga Egorsheva, CEO, and co-founder of Lobster – ’Uber’ for stock imagery, enabling anyone on social media to license and sell their photography and Dr. Aygul Zagidullina, Director of Social Media at Motoworld, – the World’s Fastest Business Translation Platform, Aygul is also Google’s Developer Group and Women Techmakers London Lead.
We kicked off the panel with a discussion around General Data Protection Regulation (GDPR), and it’s effect. Olga noted; that recommendations from the parliament are still in progress and 60% of companies will not be 100% GDPR compliant immediately. Decisions such as Facebook disabling private Instagram accounts from being available (e.g., for Lobster to list photos) can have significant impacts on small businesses.
Similarly, Antoine noted it’s difficult to fathom what GDPR will mean, although the law is there, nobody knows how it’s going to be implemented and applied. Some new rules for businesses include that you must have a Data Protection Officer and that you as a company can be liable for any soft processes that you use, e.g., if Amazon hosts your data and the break the law, your company can be responsible. Although the law is there, it’s hard to tell what will happen in the future.
The House of Lords released an AI report, and we asked the panel their thoughts on these principles as a business.
The panel and our audience were engaged and triggered by informative perspectives on critical areas of thought summarised below.
When it comes to automating the decision-making process, Aygul highlighted the importance of knowing whether we are speaking with a robot or human and having the right to request and understand, how decisions have been made. In future, Google plans to add identification showing when a machine is used.
Olga noted that algorithms are more of an ethical issue of how you train them which is becoming more and more important in healthcare, recruitment, insurance and anywhere algorithms make decisions about humans and their destiny. We have to be responsible when teaching AI as if we are teaching 10,000 children. If showing families, show a whole range of family types. When showing music, show a range of different genres.
Algorithms should not be making the final decision, and the bigger question is at what point machines should get involved.
Similarly, Antoine said it’s not either or when it comes to making decisions, and we shouldn’t have to decide between algorithms vs. humans. Humans should be using algorithms to do their work faster and efficiently, but should still make the final decision. For example in healthcare, analysing patients still requires experience and the intuition of a human.
Transport for London is praised for making their data freely available resulting in the development of apps such as CityMapper. DeepMind’s collaboration with the NHS Royal Free Trust developed an app used to detect acute kidney injuries. The panel was quick to agree that the availability of public data is a positive thing.
Olga provided the example of Toyota’s latest ad where which was personalised based on the user’s Facebook profile. Features of the car such as it save CO2 emissions or highlighting its technical features were shown based on what interested the user. This personalisation represents a good experience based on data and is one of the benefits of AI in the advertising industry. She also said she wouldn’t mind sharing in-depth data about herself if it was beneficial, for example, a juice being recommended for her in the morning based on her genes.
Aygul noted that if she shares her data, she very much wants something in return.
Social media has played an essential part in creating our digital selves. This has had an impact on the way we interact with each other; in the workplace, with our friends and families and how we relate to the world and our societies.
In Google’s recent I/O keynote, the CEO spoke about social media and highlighted the points below:
Can AI give us more time to be more human?
Olga noted that technology optimises our jobs, frees time, enables communication with our kids who are abroad, allows children to communicate with friends and social media. Antoine said that having a smartphone is like having knowledge of the world in your pocket, but could also mean constantly being on your phone and ignoring the world around you, so there needs to be a balance. Aygul praised technology for giving her the tools to work from home as a new mum. E.g., being time-poor pre and post-childbirth, AI had helped her prioritise her communication, e.g., Slack helped her pick the most important messages to read.
All panelists agreed that whatever we share on social media should be taken as public information, and there is no threat in sharing data as long as you believe the government is in good faith.
Most questions from the audience raised concerns about AI taking people’s jobs.
Antoine noted that technology changes the workplace, this has been the case for hundreds of years and AI is the new wave of this. What’s important is how does the government implement changes or help people adapt to the modern world, and teach the skills for new jobs.
Olga agreed that the government needs to teach people how to work with AI. There is a new level of skills that we can be empowered with and must acquire to stay on top of the workforce in most of the fields.
Previously revolutions have left other things for humans to do instead. When machines got better than us in muscle power, we developed our brains. When computers got better at calculating, we got to do more creative and compassionate stuff. In a few decades, what if there are no creative jobs left? And can AI bring the same fulfillment that occurs when humans are creative?
Aygul explained that a developers roles today are already progressing, e.g., coding Amazon Alexa skills. Antoine noted that all AI systems are trained on human interactions, and the beauty of art is that, incredible art is no way linked to what has been done before, it’s something new, and it will be difficult for machines to replicate that. A lot of jobs will be automated to a certain degree of creativity achieved but where will we be in 20 years time assuming that AI is applied everywhere? Nobody knows.
What challenges do we think AI could help us resolve?
Antoine says driverless cars will have a significant impact on the way we work and increase productivity on our commute. But of course, this means drivers will lose jobs.
Many applications that will have a significant impact on their respective industries for example healthcare using AI to create drugs.
How all this will unfold in the future is unknown and is an ever-evolving discussion. Our next event in September will continue this conversation around the future of work and what AI means for creativity.
Stay tuned for updates.