top of page

Dana’s Three Predictions for AI in 2019

Each year, we like to make some predictions around technology based current trends. With artificial intelligence (AI) being such a hot topic these days, we decided to dig through some of our notes and come up with our predictions for AI in 2019 within three key areas.


AI will Continue to be Democratized and Commodified

The implementation and usage of AI was once seemingly reserved for PhD’s who specialized in the black art. But today’s AI offerings consist of countless pre-built tools, libraries, cloud-based services, and a plethora of informational websites and YouTube videos, that are enabling developers and techies at all levels to add AI to their solutions.


Based on current trends, AI is will continue to find its way into new applications and will be used by more users than ever, whether they are aware of its presence or not. So, in 2019, look for even more pre-built AI offerings, that provide even simpler and less technical interfaces, that democratize AI across a broader base of users including non-developer personas. And as this progresses, we predict there will also be some big breakthroughs in generalized AI.


Consumers will Start to Demand More as AI Better Infiltrates Their Lives

According to this article on Inc.com, it’s projected that 66.6% of Americans will use voice recognition in 2019. This is almost certainly being fueled by new consumer-oriented, user-friendly devices such as Alexa, Google Assistance, and Siri, whose accuracy and capabilities are improving all the time. This is turn, is leading to a better understanding by consumers of just how friendly and useful AI can be in our daily lives.


As AI gains more acceptance amongst consumers and becomes more integrated with other systems such as home automation, look for consumers’ expectations to increase as well. While it’s amusing today, to ask a virtual assistant for a weather forecast or the time of day, we expect that in 2019, consumers will not only want, but will expect even more complex and personalized experiences based around a variety of factors such as voice tones, behavioral patterns, and gestures. This of course, should provide a hint at just how big the opportunity could be for developers here.


Similarly, as consumers start to experience just how well AI can operate at home, they will come to expect the same level of quality everywhere. For example, when communicating with an organization’s chatbots, consumers will expect incredibly intelligent interactions that lead to real-world, useful information and answers.


Watch for Increased Emphasis on Transparency and Trust

Traditionally, software algorithms have been deterministic, but AI is fundamentally non-deterministic due to its the data-driven nature and self-teaching elements such as machine learning models. So, as the use of AI becomes more democratized, commodified, and applied in new ways, so too will the demand for transparency around just exactly what is going on behind the scenes in an AI system.


Advanced users such as enterprises, will demand a better understanding of how their AI systems work and how those systems go about making decisions for us, especially when the stakes are high (think autonomous technology here). Gone are the days where a team of data scientists, programmers, and/or third parties could develop “black-box” AI systems which users could rely on solely based on past “good output”.


Going forward, enterprises will likely demand that those black boxes be made into clear boxes, to facilitate an understanding of how the machine derived decisions from the data. This will be used to both optimize the outcome, and to uncover the underlying and intrinsic information that humans may be incapable of recognizing without the aid of machines. Thus clear-box AI will not only lead to decision making output, but an additional understanding about how to make decisions, which can be even more valuable than the direct output.


Conclusion

It’s always fun to try and predict the future, especially in an area that’s as hot as AI is right now. We think that the three areas of predictions above illustrate where AI is headed in 2019, and look forward to seeing how things materialize when we look back on things next year. Now, if only AI could predict the future.


Comentários


Os comentários foram desativados.
bottom of page