5 Rising AI Startups You Should Know

Developing AI is no longer an activity that exclusively favors technology’s titans. With the rise of the blockchain and related technologies, just about anyone can get involved in the AI space, provided that their team has the expertise needed to do so. Considering this, we’ve developed a short list for you below of five companies to watch as small AI firms become more important to the evolution of different industries.

Tempus- What currently stands in the way of Precision Medicine becoming a reality? Medical data is currently heavily siloed, based on where patients are living, where companies are operating and where the relevant government’s regulations allow the data to be held. Companies like Tempus aim to put medical data back into the hands of the patient. In doing so, they allow him or her to authorize its’ usage, in order to provide the best treatment possible. If this seems confusing, that’s because it is. A slew of problems come to mind with such an idea, not the least of which is that it will be close to impossible to get patients to buy into a new form of storing and processing their medical data. Still, Tempus boasts that they already have the largest amount of medical data in one place. If this proves to be validated as true by significant partners, then it is possible that Tempus will become the data repository that serves as the foundation for future AI-driven assistants in the medical profession.

2. JukeDeck- JukeDeck offers an AI-based service that helps you create “royalty-free” music for any sort of purpose, though they advertise it as specifically catering to content creators. If queried on where the chief value of JukeDeck lies, the company itself would argue that their musical compositions rival even those of experienced human composers in their uniqueness. Though not a lot of detailed information seems to have been made public on the subject, it is clear that the company offers on-demand “generative” music, which appears to refer to JukeDeck’s system’s ability to take in data from human-generated compositions to learn how to be a composer. On the other hand, since their main system appears to be based on principles related to supervised learning, they likely have not taken in enough labeled data to truly optimize its’ performance. To make matters more convoluted, they’ve been criticized for taking a fee for their service even though it’s not exactly ready, though they do offer a free option. As time goes on, it will be interesting how their work with StructureNet, which is the Recurrent Neural Network that appears to drive JukeDeck’s compositions, plays into its’ chances of market success. In the future, I’ll dig into what Recurrent Neural Networks are, but for now, keep in mind that they’re the same sort of AI system that drives Siri, of Apple fame. What makes them unique is that they always consider both past and present data at the same time(or consistently backpropagate). Arguably, in doing this, they can make more informed predictions than other types of neural networks, which at least begins to explain their popularity.

3. FiveAI- Time and time again, it’s been suggested that only the leaders of the technology and automotive industries will be able to compete in the vertical of self-driving cars. FiveAI is one of several small companies that is looking to prove this prediction wrong. Unlike most of this group, they’re choosing not to target the consumer vehicle space, in favor of getting involved in autonomous public transport. Currently, they exclusively work out of the United Kingdom, though they make it clear on their website that their ultimate goal is to develop something that targets the entire European continent. Over the course of their short history, they’ve run two pilots of their self-driving cars on the streets of London. It’s important to note, however, that neither of these pilots involved passengers and the latest only involved five vehicles. Therefore, everything they’re doing at this time is quite early-stage. Still, in 2020, FiveAI plans to finally put passengers in their vehicles on a small-scale. Perhaps, if all goes well, they’ll be the first company to debut the live-usage of AVs on the streets without any sort of recorded controversy. To be fair though, the cards seem to be stacked against them. Since the success of any sort of fully autonomous AI agent largely depends on unsupervised learning frameworks, the time in which humans aren’t driving anymore is a ways off, to say the least.

4. Shield AI- Shield AI is the only company on this list that is involved in the global national security sector. What they specialize in is creating various sorts of drone-like robots that can work in the air, on the ground, or underground to report back on “dangerous environments.” Since the company’s mission explicitly states that it is aiming to “protect” both those who protect civilians and civilians themselves, it is reasonable to expect that their offering will have military-related applications. Because of this, it will be interesting to see if Shield AI can achieve its’ ambitious vision of providing benign, yet protective services for the global populace. Standing in their way is the commonly held fear of a Skynet-like scenario where humanity loses control of the AI systems that they created. Still, if the “age of Artificial General Intelligence” is actually as far off as it seems to be, Shield AI’s team will have plenty of time to figure out how best to address these concerns. If you’re interested in their actual products, check out their “Nova” drone, which is reportedly capable of flying missions autonomously after being trained on the relevant routes. Since all of their products will be connected to what they call their “HiveMind,” it is logical to conclude that any chance of their success will depend its’ success. This “Hivemind” is a massive neural network that according to the Shield AI team, holds millions of data points, especially about what they call “GPS-denied areas.” Because of this, it will also be truly interesting to see exactly what sort of missions the Hivemind ends up facilitating in the future.

5. Skydio- By now, you’ve heard of autonomous vehicles time and time again from different sources. At the same time, most publications still almost exclusively focus on self-driving cars and nothing else. With that in mind, imagine the implications if drones were made fully autonomous, which is Skydio’s primary goal. Given that it has been claimed that their drones can already fly autonomously with no human assistance, it’s only a matter of time before AI detractors start decrying the inevitable onset of Skynet. In reality, however, it appears that Skydio is catering less to institutional customers and more to the individual consumer. Their drones are marketed as if they are the inevitable next iteration of the camera, i.e., what it should evolve into after smartphone cameras are made obsolete. Still, since their charging over $1000 for these drones, it would appear that this particular milestone has not been crossed just yet. In the end, like Tesla, they may be slowly working their way towards more affordable options, while evolving their product offering with the help of their early adopters. For now, imagine Skydio’s drones as the next cool tool of successful content creators.

This list is only a tiny cross-section of all of the AI startups that currently exist. In future posts, we’ll analyze some of these companies to examine further why they might make a unique impact with their technologies. Until then, if you’re interested in digging into any of these topics further, check out some of our suggested resources below.