2018 was a huge year for technology, with many firsts and even more advances. An AI-generated painting sold for six figures at Christie’s, Google’s Duplex assistant held convincing conversations and augmented videos used the superimposed faces of celebrities for “deep fake news”.
Amazon’s Alexa became commonplace in the households of not just millennials and tech’s early adopters, but also those who were traditionally technophobes. Meanwhile, self-driving cars became a reality – even for those who weren’t driving them, their presence in international news meant AI was a talking point for families around the world.
One thing is for sure, there will be many more grand claims for AI’s capability to change the world as we know it. We will see new industry statistics published and market forecasts released which will undoubtably bring AI even further into the spotlight for the year ahead.
So, what can we expect to see from AI in 2019?
In 2019, we can expect to see continued growth in investment into AI initiatives and research. Key milestones in the UK last year included a significant investment in AI, technology and digital services as part of the Budget, and a promise to introduce five new AI medical technology centres in 2019. There is a global arms race to become the world’s most advanced tech innovator as nations engage in a technology tug-o-war, and the UK is vying to be a key player.
As it becomes more commonplace, society’s understanding of AI will improve with perception becoming more favourable. AI will continue to deliver on its promises in both the consumer and B2B markets and its benefits will become increasingly apparent, accelerating further adoption across all markets.
In 2018, we saw rapid advances in the self-driving car market, and traditional carmakers such as Audi pledged to invest in this technology. As driverless cars continue to grow ever-closer to production, pressure will mount on countries to soften regulation and encourage innovation in this field.
Autonomous vehicles offer a huge advantage compared to traditional vehicles in terms of their safety and efficiency and, like with most technology, it needs to be seen to be believed. Vast numbers of accidents and deaths are a result of human error and preventable, and this technology could be huge in eliminating this.
Continued growth of open-source AI projects will democratise AI and put the power in the hands of the masses. Facebook has already announced Horizon, an open source toolkit that extends the applicability of ‘reinforcement learning’, and I predict that other huge technology companies will follow suit.
Life as a service
AI will accelerate the proliferation of ‘life as a service’ models where lifestyle needs are delivered on demand or on subscription. Consumers will move away from owning products, instead gravitating toward subscription models with enhanced predictive capabilities, such as shopping assistants that anticipate food preferences, automatically order and have the ingredients ready and waiting to go. Early iterations of these services include Apple Music’s predictive playlist functionality were facilitated by AI.
We will see a growth in the number of large-scale enterprises building AI products and programs in-house for both internal and external use. More corporates are looking to hire tech talent in-house and build bespoke platforms and solutions, rather than outsourcing off-the-shelf solutions widely available. The ROI of these projects is set to determine how the future landscape evolves, but I believe the employment opportunities and market demand for developers and data scientists will continue to increase.
As consumers become more conscious about protection of their data, and in a GPDR-context, privacy will be at the centre of innovations in 2019. Developers who embrace the opportunity to use AI to overcome these concerns are more likely to succeed at scale, or face scrutiny by users and lawyers alike.
2019 will be the year of separating the wheat from the chaff when it comes to using AI as a ‘secret sauce’. If in practice the technology does not perform to expected levels or is an imposter of other, more functional services, it will not survive in the market. Consumers are waking up to what AI really means and are taking a more informed view of product ROI.
Reinforcement learning for AI software is set to see greater adoption in 2019, according to our research partner Tractica. Already evident in manufacturing and energy, we can expect to see a push to use unsupervised or semi-supervised machine learning models which will offer a realistic method for unleashing AI’s potential.
As AI becomes more commonplace, businesses will be required to adopt a far more sophisticated strategy that incorporates rigorous planning, research and development. Business leaders will be expected to focus on harmonising collaboration between human workforces and technology as well as ensuring that AI is responsible, trustworthy and ethical in all dimensions.
2019 is set to be a fascinating year for innovation and mainstream applications of artificial intelligence as worldwide AI spending is set to grow from $8.1bn to $12.7bn in 2019. I look forward to seeing how my predictions fare.