In 2020, Synced has covered a lot of memorable moments in the AI community. Such as the current situation of women in AI, the born of GPT-3,AI fight against covid-19, hot debates around AI bias, MT-DNN surpasses human baselines on GLUE, AlphaFold Cracked a 50-Year-Old Biology Challenge and so on. To close the chapter of 2020 and look forward to 2021, we are introducing a year-end special issue following Synced’s tradition to look back at current AI achievements and explore the possible trend of future AI with leading AI experts. Here, we invite Mr. Sheldon Fernandez to share his insights about the current development and future trends of artificial intelligence.
About Mr. Sheldon Fernandez
Sheldon Fernandez is CEO of DarwinAI, the explainable AI company that enables enterprises to build AI they can trust. The company was named a cool vendor in Gartner’s October 2019 Cool Vendors in Enterprise AI Governance and Ethical Response report. CB Insights also selected DarwinAI for its AI 100, CB Insights’ annual list of the 100 most promising private AI companies in the world. Based on years of distinguished scholarship, the company’s patented explainability technology accelerates advanced deep learning design. To learn more about DarwinAI, visit their website at www.darwinai.com or follow them @DarwinAI on Twitter.
The Best AI Technology Developed in the Past 3 to 5 Years: “Attention Mechanisms”
Attention mechanisms in deep learning. Attention mechanisms are unique, stand-alone architectures that allow a deep learning model to focus on ‘what’s important’, facilitating more effective and trustworthy decisions and increasing their representational power.
The impact of this breakthrough is most evident in the field of natural language understanding, where the introduction of the Transformer self-attention network architecture has completely revamped the way we design language models (as exemplified by GPT-3), and which is now being applied to impressive effect in computer vision.
The Most Promising AI Technology in the Next 1 to 3 Years: “Explainable Artificial Intelligence”
Explainable Artificial Intelligence (XAI). As AI increases in scope and complexity, there is a growing demand for fairness, transparency and ethical norms in regards to their design and execution. The ‘black box’ problem that plagues AI – our inability to peek inside neural networks and understand how they work – represents one of the most critical moral and business imperatives of our time. Such urgency is evident with policy developments in North America and the EU that are mandating increasingly stricter forms of algorithmic transparency.
The two layers of XAI: ‘how and why’ a system makes a decision as well as when and where it can be trusted (or not trusted), will be in key in allowing designers to build robust AI solutions that meet regulatory and compliance requirements for the public trust.
The Biggest Challenge in the Field of AI: “Bottleneck in Understanding”
While vast amounts of energy have been expended improving the accuracy and performance of AI systems, a significant challenge is understanding how these systems achieve such levels of performance. Specifically, the growing complexity of deep neural networks is mirrored by our inability to understand the behavior and drivers behind their predictions.
This bottleneck in understanding has started to impede advancements in AI in two ways.
First, it is becoming increasingly challenging to devise new ways to improve such systems (new architectures? new training strategies? more data?). Second, it is difficult to determine if such systems are making the right decisions for the right reasons instead of relying on spurious correlations and erroneous cues.
Both academia and industry have begun to tackle these challenges through advancements in Explainable AI, but much work remains to be done.
The Latest Noteworthy Development: “DeepMind’s Recent Achievement Around Protein Folding”
DeepMind’s recent achievement around protein folding should be widely celebrated and talked about. Beyond its thematic connection to COVID and vaccines, the organization’s progress in this area is of tremendous relevance to drug discovery, computational biology, and pharmaceutical chemistry. It is a compelling example of AI advancing the pure sciences which hopefully prefigures additional work and cooperation in these spheres.
With all our focus on Artificial Intelligence, it is easy to lose site of the ‘Biological Intelligence’ that make such innovations possible. So as we continue to tout the evolving sophistication of AI, let us not forget the most awesome device in the cosmos – the human brain – that’s behind it. In the words of English biologist Thomas Huxley:
“How is that anything so remarkable as a state of consciousness comes about as a result of irritating nervous tissue, is just as unaccountable as the appearance of the Djinn when Aladdin rubbed his lamp”.
This report offers a look at how China has leveraged artificial intelligence technologies in the battle against COVID-19. It is also available on Amazon Kindle. Along with this report, we also introduced a database covering additional 1428 artificial intelligence solutions from 12 pandemic scenarios.
Click here to find more reports from us.
We know you don’t want to miss any news or research breakthroughs. Subscribe to our popular newsletter Synced Global AI Weekly to get weekly AI updates.