Rise of AI Berlin 2019 — Conference Recap
We had another fantastic meetup last week at the 2019 Rise of AI conference in Berlin. It was a packed house with top speakers sharing their best insights on the current and future landscape of Artificial Intelligence.
This truly is Europe’s leading conference on Artificial Intelligence. If you’re looking to understand AI, learn about its applications, and be inspired by the future of AI, the Rise of AI conference is the place to be.
Here are some of our main takeaways from this awe-inspiring conference.
Biological research that inspires technology
Prof. Dr. Tim Landgraf kicked things off with a curious and entertaining talk on bees and what they can teach us about learning — about how modern biological research brings new challenges and inspiration to technology.
He’s known for building the first working honeybee robot which was able to communicate with nestmates via the bee dance, and fish robots that integrate into live fish schools. He has even built a system to track all bees in a colony over their lifetimes to understand how information is processed by the bees’ social network.
Their projects are mostly biologically focused, but, in using machine learning, they have developed a new type of generative adversarial network, RenderGAN, to generate images of honeybee markers with semantic control over the labels, or low-noise embedded amplifiers to record brain signals from bees on a copter.
Dr. Landgraf has conducted amazing studies. But our takeaway goes beyond that.
What dazzled us is how in using AI to seek optimization and efficiency, his team created a totally new technology that has other applications.
In his fascination for both the individual and collective intelligence of bees, Dr. Landgraf envisions future scenarios. One of which is where cars recharge other cars, much like bees feed other bees to maintain high energy levels and greater output.
A revolutionary technology that will open new doors to the market of small rechargeable items for small pockets and that will transform our perception of road congestion.
For an in-depth explanation of his AI-driven application of bee-smarts to human “swarms” and self-driving cars, check out this TED talk.
Faster processing and fraud detection
On the topic of optimization, Berlin-based startup Omni:US is using AI technology to transform insurance companies from process-driven to data-driven.
Every day, insurance companies deal with vast amounts of highly valuable documents, which they need to process with speed and accuracy.
AI and machine learning can help revolutionize these workflows by reading, understanding, and extracting relevant information and automating time-consuming tasks such as claims management — completing in minutes what may have previously taken hours, days, or even weeks to do.
The AI also becomes more intelligent and is able to learn based on increased exposure to data/information, which allows the system to continuously improve. The system can even read and process handwritten information!
AI in insurance can also decrease the risk of fraud. Claims settlement managed by employees comes with the risk of human error — it can be easy to overlook possibly fraudulent information and lose a payout.
AI, however, can detect the misrepresentation of information by carrying out tests to validate a claim. This process involves cross-referencing public information against the metadata supplied by claims documents — a job that may prove intensive and painstaking to a human — but not for AI-powered technology.
Meeting Bert
Jakob Uszkoreit (of Google Brain) talked to us about BERT, or Bidirectional Encoder Representations from Transformer. It’s a human language analysis neural net model that has proven to be exceptionally good at answering questions.
A pleasant surprise — Google has open sourced the model so that everyone can use the code.
Natural Language Processing, or NLP, lays at the cornerstone in the latest developments in AI. NLP is a tool for computers to communicate with humans effectively, including different components like speech recognition, sentiment analysis, and machine translation.
NLP brings computers closer to a human-level understanding of language. As you can imagine, computers don’t intuitively understand the way humans “do”: Understanding even the most simple phrases can prove to be very difficult.
Google’s BERT is a significant leap in the world of Artificial Intelligence. A computer system like this can learn, understand, and process the small idiosyncrasies of language and use them to tackle specific tasks.
Though BERT is a big leap in the right direction, researchers are nowhere near where they want to be with AI, and BERT is still far from having a human-level of common sense. Who knows, in the near future, you could have late night conversations with your home system. BERT is currently already trained in 102 languages.
Empathic AI in Healthcare
Three years ago, Spanish telecommunications giant Telefónica established Alpha, a lab in Barcelona working (in stealth) on innovative technology that holds the promise of potential new revenue streams.
Today, they’re exploring new approaches to machine learning in the hope of building empathic, ethical software.
To keep reading and access our comprehensive article, be sure to head over to our blog.
Remember, AI technology is growing exponentially and the competition out there is fierce! Startups must always prepare for the future and protect their intellectual property.
We are your go-to resource for all things Patent and IP! Learn how you can secure and grow your startup with instructive daily insights from our Learning Hub: https://kisspatent.com/learning-hub
Still hungry for more? There is a wealth of information in our Resource Center! Explore our ebooks, whitepapers, webinars, and more: https://kisspatent.com/resourcecenter
If you have questions regarding your particular business scenario, don’t hesitate to reach out. We’re more than happy to help! Schedule a call with our experts here: https://kisspatent.com/contact
Originally published at https://kisspatent.com on May 23, 2019.
