CB Insights have undertaken the review using their NExTT framework, to examine a range of emerging artificial intelligence trends. The analysis additionally considers industry adoption and market strength of each trend. The report also categorizes them as necessary, experimental, threatening, or transitory (these factors spell out the NExTT acronym).
The report is titled “AI Trends In 2019” and each area reviewed is structured around the categorization factors. These are defined as:
Necessary: this means fast-track regulatory approval by health authorities for products like IDx-DR. This tracking is opening up new commercial pathways for AI imaging and diagnostics companies.
An example is open-source software, which has enabled the barrier to entry in AI to be lower than ever before. A leading innovator here is Google’s open-sourced TensorFlow. This platform makes AI accessible to everyone, and companies like Google, in turn, benefit from a community of contributors helping accelerate its AI research.
Experimental: As an example, Apple is changing how data flows in healthcare and is opening up new possibilities for AI, specifically around how clinical study researchers recruit and monitor patients.
Another example is early-stage research that combines biology, physics, and machine learning to tackle one of the hardest problems in prosthetics: dexterity. One compony in this field is DARPA, which is investing in an advanced prosthetics program.
The concept of ‘edge AI’ also falls in this space. This means moving away from cloud-based AI, to AI systems that can take decisions in real time based on collected data at ‘the edge’ of what ever the AI is interfacing with. In this space, Nvidia, Qualcomm, and Apple, plus several startups, are focused on building chips exclusively for AI workloads operating at the “edge.”
Transitory: With this, major pharmaceutical companies like Pfizer and Novartis tapping into AI SaaS startups for innovative solutions to the long drug discovery cycle.
Threatening: This category is about researchers are pushing boundaries with reinforcement learning, but the need for massive data sets currently limits practical applications. So far, reinforcement learning has particularly taken off in gaming and robotic simulation. However, to expand into other fields, greater access to data is required.
The NExTT framework is intended to help educate businesses about emerging trends and to guide them with their decisions in accordance with their own level of comfort and business risk.
