Artificial Intelligence in 2019: 15 Predictions for the Social Impact of AI

2019 will be the year we’ll see the democratisation of data science – creating a profound and lasting impact on the way it is used around the world. This will be driven primarily by open source tools enabling more people to begin contributing valuable insight and analysis to data sets without having to get to grips with proprietary systems.

It will enable organisations to begin unlocking the power of their data through technology as they also benefit from accessing the widest possible pool of talent, no longer restricted to those who can use highly specialised software.

Remember the days when voice assistants were novel? Voice computing is permeating everything we do, from home assistants to hands-free driving, to scheduling your appointments in conversations with other people.

The days of keyboard and screen-only machine-human interactions are over. It’s not just voice; our tech is wearable, tactile and augmented. These human interactions present not just a new way of interacting with technology, but whole new streams of human data that can present new opportunities, such as diagnosing neurological conditions from voice patterns. Enterprises will need to tap into this data to dramatically improve their digital services and customer care.

We’ll continue to see a rise in technology created to interact with humans not through a screen, taking personalisation and automation to whole new levels.

Ethical concerns around the use of data and technology have risen sharply after the last year and will continue to do so. In particular, the speed at which AI is evolving is raising questions, highlighted by the recent announcement that Canada and France are joining together to establish a panel to police unethical uses of AI.

There has been a spotlight on the possible negative effects surrounding this technology rather than AI’s potential as a force for good. As with many technological advancements, it’s the mindset towards this technology that needs to be addressed, rather than the technology itself.

To tackle this, companies such as Microsoft and IBM are making ethical thinking a focus, creating courses that provide organisations with the latest insight into ethical development. As an increasing number of businesses create more ethical strategies, AI will be more confidently adopted, and developers will be empowered to break new ground

In 2019 we can expect conversations around AI and tech ethics to intensify, especially when it comes to consumer data and privacy. We may also see the field widen its gaze to less mature technologies, like augmented reality.

Importantly, more companies will be trying to anticipate how their tech could negatively affect us in the future and making efforts to bake ethical principles into product development as standard. Those with new mass market products launching should expect to be scrutiniSed by the public, politicians, and the media. Those developing the next big AI hit should be thinking about future-proofing.

In 2019, users will start getting rewards for good data that helps improve AI models. Online services like Facebook and Google collect our personal data on a daily basis, which is then transformed into large advertising revenues.

However, users are not receiving any benefits because these sites are free, and instead, the users become the product themselves as they continue to benefit these companies. To change this, users will get rewards starting with small upstarts and interesting projects using properties of cryptocurrencies to share user-data monetisation with users.

Even though not initially using AI methods, this is similar to what we hope to achieve with our own technologies; paying users for the quality of content they create. Eventually when the volume of this data hits a critical mass, we too hope to be able to use AI models to measure data and content quality.

We all expect machine learning to continue to advance and revolutionise almost every aspect of our lives (whether we are aware of it or not).

Beyond this I think convergence will be the key trend for 2019. Currently both established and emergent mediums act in a somewhat isolated fashion. Voice computing, messaging, web, mobile, VR and AR are all somewhat separated from each other by artificial barriers at the moment.

We are already seeing signs of progress here - in hardware with devices like the Google Home Hub, Echo Show / Spot, Magic Leap One and in software with standardisation initiatives enabling, for example, increased native features for PWAs and cross browser and device AR implementations.

I would expect mediums to blur significantly in 2019, with communication between and within different mediums becoming much more seamless. The emergence of further voice powered devices with screens and mixed reality headsets should providing major impetus here. I would equally expect a plethora of innovative services to arise to enable creation in this newly convergent world.

The biggest impact for everyday people will be the growth of voice and screens. A combination of smart speakers with screens, digital assistants on mobile devices and data-informed recommendations will enable brands and retailers to reassure shoppers and create useful new brand experiences, whilst providing personalised and guided options for maximum impact.

AI will continue to augment digital advertising and empower marketers to make better decision with their ad spend. This year, we saw a huge growth in spending on digital ads, with £6.36bn invested in the first half of 2018 alone, and this will only continue to grow as more and more brands incorporate AI into their arsenal of digital advertising tools. We will see it being used more to ascertain whether the data gained from online channels is being utilised effectively; this is where AI will differentiate campaigns as it will enable marketers to decipher audience engagement and create personalised ads that truly resonate.

We also expect AI to all but eliminate most brand safety concerns in 2019 and mark the ‘safest’ year on record for brands. 2018 was a tough year when it came to brand safety; Cisco and Mars, for instance, removed their ads from YouTube due to safety concerns. However, with AI becoming increasingly sophisticated, it can help brands serve safer and more secure ads, and create a more transparent industry as a result.

And finally, AI will create a much more personalised advertising world. Personalisation is a term that has been cited in advertising for years, but until now, the capabilities have been limited to consumers’ previous preferences, purchases or searches. Now, thanks to AI, marketers can make real-time decisions on when to reach the right consumer, with the right content, at the right time and on the right platform. It’s exactly the kind of engagement consumers expect, and only AI can deliver it and it is something we envisage brands and advertisers will have to take advantage of in 2019.

Article 22 of the General Data Protection Regulation (GDPR) states that customers have a right to understand why a model has made the decisions it has about them.

Despite this, the immaturity of the AI industry overall is painfully evident when the conversation turns to the explainability of machine learning algorithms. Questions about the details of learned patterns in neural networks are usually answered with quizzical silence, even at the largest companies.

This situation is completely unacceptable for anyone who has to explain decisions taken by an AI model to a customer - “computer says no” simply won’t be enough in 2019. Hence, we can expect more transparency to be built into AI models.

2019 will see rising conflict between the privacy concerns of citizens and the goals of AI developers. There will be greater scrutiny on how personal data is processed within AI systems and a higher burden on AI developers to explain how algorithms work to both consumers and regulators.

The key catalysts for this conflict are the multiple parliamentary investigations into fake news that are placing the transparency of social media companies and their AI algorithms in the spotlight.

In countries such as UK, USA, Canada, India, Singapore and France the inadequacy of social media platforms' response to investigation have sparked calls for new ‘digital bills of rights’ which would strengthen property rights in personal information, biometrics and face images, with a focus on regulating AI at their cores. Many of these investigations are expected to escalate and conclude in 2019.

As societies confront the worst cases of AI misuse in 2019, AI developers must be prepared to explain and defend their data practices.