2019 is approaching, that means that it is time for some new predictions. We decided to ask AI at Work guests three questions. Don’t miss their answers!
Is 2019 the year AI adoption at work crosses the chasm and becomes mainstream or are we still a year or two away?
Erin Winick, Associate Editor, Future of Work, MIT Tech. Review: “It depends on what you mean by mainstream. We already have AI in so many of our everyday products. I think we are still a few years away from it becoming a more democratized and accessible tool, but I think next year it will continue to enter more parts of our lives, even if people don’t realize it.”
Steven Peltzman, Chief Business Technology Officer, Forrester Research: “I think it could be; though my guess is that while the technology might be ready, it will (ironically) be the humans that hold it back. They'll hold it back by not understanding how to find the right fits between AI and their business & customer engagement models, and they'll underestimate the work needed to train the AI and deploy it properly, especially when it comes to the end customer (or even employee) experience. So I think there's going to be another year or so of companies stumbling and learning from those experiences.”
Jim O’Neill, Former Chief People Officer, HubSpot: “I would say 2020 is more likely. I think there will be more breakthroughs in 2019 but nothing that is broadly available to make an impact on mainstream society.”
Matt Scott, Co-Founder and CTO, Malong Technologies: “I absolutely expect a few – but only a few – large enterprises with a history of innovation leadership to jump from the pilot projects and experiments they’re trying today, to full-scale implementation of AI in 2019. For fast followers and other occupants along the balance of the traditional adoption curve, they’ll watch and learn, and make the investment in the ‘20s.”
Enrico Palmerino, CEO, Botkeeper: “We are still 3 years away. I think AI is functionally or industry limited such that though maybe more than 50% of companies leverage it in one way or another that only. 5% of the existing workforce actually interacts or personally leverages AI and it's not until a majority of people are using AI in their jobs that I would consider it crossing the chasm.”
Rob May, CEO, Talla: “Possibly. I think if a recession hits, companies will look for ways to increase productivity and save money, and AI is a very powerful technology for that.”
What AI related technology or issue are you most excited about in 2019?
Erin Winick: “I’m very excited about the collaboration between AI and robotics. Robots have continued to get more dexterous, even breaking records in the past year. I’m looking forward to seeing how these technologies advance together and am interested to see the implications these advances have for the manufacturing floor and for warehouses.”
Steven Peltzman: “Selfishly, I'm ready for AI to start delivering a significant improvement to my daily life, even if in small ways. I want websites and apps to delight me with content, goods, services, and information I didn't even realize I wanted. I want my bank to figure out on its own thru my app that I'm on vacation, and expect to see credit card charges from that place without freezing my account. I want my customer & healthcare experiences to be smoother, easier, and faster with AI. Instead of presenting me with "Alice", or "Julie", or "Siri", I'm excited to see companies of all kinds finally move past empty, flashy AI and use it to delight and serve us all in meaningful ways, even if those ways are small and incremental.”
Jim O’Neill: “Ability to identify ways to up-cycle existing, unused resources to reduce further production of goods. I don't know of any AI being currently applied to these types of activities (in light of a lot of AI being applied to renewable or alternative energy sources).”
Matt Scott: “New chipsets. Running on the edge is becoming the norm for many AI applications in industry, and there’s a need for lower power consumption and lower cost. As a result, efficient edge chipsets are in demand, and manufacturers are responding.”
Enrico Palmerino: “I am most excited about using AI as our knowledge base. This means that both department related questions and general company questions are answered by a bot more accurately AND consistently then people. The result being a more aligned team, a more cohesive culture, and a complete removal of minutia from our most valuable employees. Thank you Talla, you are now the go to person for all botkeeper questions.”
Rob May: “GANs... I think they have many applied uses that aren't implemented yet.”
If you are going to be an AI Grinch, what is the biggest concern you have for AI in 2019, or something you think is still holding AI back that needs to be fixed?
Erin Winick: "My biggest concern remains bias in AI. It’s such a difficult thing to spot and weed out, and as more people begin using AI, I’m worried about how biased algorithms could deny people opportunities or negatively affect their livelihoods."
Steven Peltzman: “Listening to Forrester's analysts, I agree that probably the biggest potential hurdle for AI will be the lack of quality data on which the machines can truly learn. Algorithms will come and go, but the right data and the ability to feed machines with it on a regular basis are going to be a big hurdle. In the consumer world, this could become a problem since, while people will quickly crave the value that AI can bring them, they don't yet want to trust the companies and governments with that kind of personal information (hence GDPR and regulations like it that are on the way). I wonder if people will slowly give up their privacy as the allure of AI value grows (Minority Report, anyone?!), or if perhaps blockchain or other tech can somehow resolve this conflict between privacy and AI value.”
Jim O’Neill: “A security-related/cyber-type scenario that while might be low impact, creates undue panic or attempts at regulation.”
Matt Scott: “Geopolitical rivalries. Malong is a Shenzhen-based AI company with a Chinese co-founder and an American co-founder. We’ve been saying for some time that it’s great governments are investing in national programs to support AI development, but it’s really harmful if those efforts devolve to a fragmented system of China AI vs. Canada vs. UK vs. France vs. USA and so on. Let the companies do the competing, but let’s also maintain the values of global collaboration and openness that have helped AI grow to this point.”
Enrico Palmerino: “I think there are a lot of really cool AI technologies but a surprisingly limited number of practical AI applications. I also think the challenge next year will pose is a bunch of companies claiming AI that don’t actually have AI and thus distinguishing between true AI vs great automated workflows will be the challenge. Thus, the challenge will be differentiating between what is AI vs. Just what is automated process.”
Rob May: “The field is still too focused on research rather than applications. Too much of the news is about "breakthroughs" that are impractical, only work on highly controlled synthetic data sets, and we aren't covering enough of the real applied stuff.”