How many times have we heard that the “AI revolution has come” in the last few months?
That “AI is the future” and that the companies can either embrace the full potential of data-fuelled technologies based on the omnipotence of deep learning or they will cease to exist in the wink of an eye? I guess if you’re will be interested in this report, the answer might be “Too many.”
The reality, of course, is a bit more complicated.
The adoption of AI was sped up by billions of dollars being pumped into the ecosystem. As it is with any hyper-growth, this led to several explosive victories, such as Deepmind’s Alpha Zero and UIPath’s stunning successes with robotic process automation, but it also generated some growing pains.
The overhype of AI transformed research conferences, which were once intellectual feasts for academics, into industry roadshows that sold out in minutes and began to feature celebrity performances. What used to be a venue frequented by academics from Stanford and Oxford has become a corporate fair with an enormous presence of the most valuable corporate IT brands, such as Google, Facebook, and Amazon (check out Figures 1 and 2 for more data on that).
The most important growing pains, however, are the lack of appropriate methods to interpret AI and to keep it under control (and not deepfaking Zuckerberg into spreading conspiracy theories). Both of these issues are covered in this report, prepared for you here by Tooploox and MicroscopeIT, where we try to bring things down from the overly dramatized and judge these issues based on their realities.
So while the persecutors might call AI to the stand and cry, “You promised us autonomous cars by 2020, then by 2025, and now you’re saying they may never arrive!” Our experts (acting as the defence) might answer, “Autonomous cars are tricky, but look where we got in pursuing this dream – AI is entering every household with low-powered devices and kitchen appliances, like ovens and fridges. Look at our advancements in automatic medical diagnosis, or in smart voice assistants, or in robots that could begin to understand humans better than we can understand one another. Cars are only a matter of time.”
For the truth is, the software development paradigm has shifted – data-driven methods are not just features anymore, but essential tools that show their presence across the board, from navigation apps to CRMs. AI is now a fully-fledged production-ready technology, with great ambition ahead, and with an unprecedented amount of value delivered so far.
Final verdict? The jury is unanimous, AI is here to stay and will flourish more than ever before.
Why should you read 2020 AI Trend Report and how can you use it?
This report is the collaborative work of our experts here at Tooploox and MicroscopeIT. As a group of researchers, scientists, and practitioners, we have the skills to create powerful AI platforms: and we don’t take that responsibility lightly.
We have approached this project with the intent to explore the major AI trends as we see them, in a non technical way that everyone can follow, because we believe it’s important to make complicated ideas more accessible. Every article here not only explores the social implications of the trend, but provides basic background on how the technology behind it works. Some of the trends (i.e. Edge AI) have already hit critical mass and will see mainstream adoption in 2020. Others (such as Emotion AI) are in their infancy, but with implications so large that we felt it important to highlight their potential in the here and now.
So, how can you use this trend report?
If you’re still wondering if AI advancements in your industry are ready to help your business straight away or even to become the core of your services, you will find those answers here. With each trend, we also aim to provoke a conversation. We encourage you to discuss these trends with your peers and teammates and tell us what you think on our social media.
Commentary and Introduction by
Tomasz Trzcinski, CSO and Partner at Tooploox & Assistant Professor at Warsaw University of Technology
Szymon Stoma, CEO at MicroscopeIT & Senior Scientist at ETH Zurich
Read also about Augmenting AI image recognition with partial evidence