AI Approaching: MIT Tech Conference Coming This Weekend

February 20 marks the 2016 MIT Tech Conference, hosted at the MIT Media Lab. We are thrilled to host an impressive array of professors, professionals, researchers and practitioners, all with varying perspectives on, and experiences with, Artificial Intelligence. 

To provide a preview of this weekend's event, we've included brief bios for our slated keynotes:

Ray Kurzweil – Author, inventor, futurist
Inc. magazine describes him as the "rightful heir to Thomas Edison." PBS selected him as one of the "sixteen revolutionaries who made America." In 2012, Kurzweil was appointed Director of Engineering at Google, heading up a team developing machine intelligence and natural language understanding.

Rob High - IBM Fellow, Vice President and CTO, IBM Watson 
As a key member of the Watson Leadership team, High works collaboratively with the Watson engineering, research, and development teams across IBM.

Please see PRNewswire for a press release on the 2016 MIT Tech Conference.

AI in the Enterprise: Slack hires new head of Intelligence

Slack Technologies Inc., the real-time messaging platforms for teams and corporations, recently announced the hiring of Noah Weiss, former VP of product at Foursquare. Weiss will head Slack's Search Learning and Intelligence group. Imagine the data mining that can come from all of the conversations carried across slack. Cue more predictive analytics, reminders and updates based on textual analysis -- aka, more intelligent messaging. 

Source: WSJ Blog

Marvin Minsky: The passing of an AI pioneer

While Mr. Minsky sadly passed away last Sunday in Boston, his legacy at MIT and the AI community lives on. Minsky,  a computer science educator at MIT, was an early advocate of the potential for artificial intelligence to enhance human life. Minsky co-founded the MIT Artificial Intelligence Project (now called the Artificial Intelligence Laboratory) and was influential in advancing modern robotics through his development of visual sensors and mechanical hands with tactile sensors. We remember Mr. Minsky with admiration and gratitude for the mark he made on MIT and the scientific world at large. 

Source: The MIT Tech Review (Online Edition)

Mark Zuckerberg Posts: “We should not be afraid of AI”

It is no secret that artificial intelligence has the potential to enhance human life in nearly ever sphere of our being; regarding health, logistics, etc. the potential for increased efficiency as a result of AI abound. But while AI excels at pattern recognition, machine learning falls short in common sense. Zuckerberg contrasts the process of teaching machines to recognize patterns vs. teaching machines to actually learn.

Source: Facebook

Google Pushes the Frontier of Hardware Supporting Artificial Intelligence

Bytes on "Quantum Computing": We know that all applications of algorithms that can be grouped under artificial intelligence rely on very large data sets. And one of the drivers of AI applications in recent years has been reduced cost for the hardware that makes large-scale computation possible.  AI is a different paradigm for software and computation, and many companies have been experimenting with different types of computing infrastructures. For example, IBM Watson has its own proprietary technology that powers its Watson suite that is an alternative to traditional digital computing.

Source: The Star

Chinese Company Baidu Applies AI to Help the Blind See

Bytes on DuLight: Baidu has developed a prototype of a product that applies image recognition to improve quality of life for the visually impaired. The product, named DuLight, is an earpiece that captures your surroundings via a camera and then sends that data to your smartphone. The accompanying application on your smartphone will then process that information and narrate its interpretations. So, using image recognition algorithms similar to those that are used by Google search to help you search images, Baidu can understand a DuLight user's environment and then help the user understand that output in real-time.

Source: WIRED

MIT's CSAIL Trains Algorithm to Rate "Memorability" of Your Selfie

At first read, dedicating CSAIL researchers' time to rating the memorability of a selfie does not sound like a productive use of their time. However, what the researchers have achieved has far wider applications. Right now, many "memorability" algorithms exist. However, the one that Aditya Khosla and his colleagues have created performs 30 percent than what's currently available. The bar of performance that algorithms have to meet is set by actual humans who participate in studies, organized by researchers, in which they are asked to rate memorability of several images shown by researchers.

Source: TechCrunch

Bytes on Visual Processing and "Memorability": As researchers are able to better understand how to better mimic the way that the human brain processes and reacts to visual information, businesses can apply the lessons learned to many different use cases. A few examples include:

  • Online teaching content can be created to optimize for memorability and thus, effectiveness
  • Health and wellness-focused applications can create content and interfaces that have the highest chances of being remembered by their users
  • Healthcare providers can leverage the knowledge to screen for and identify memory-related illnesses
  • Commercial digital artists that create work in any visual medium can create more evocative content
  • And, of course, more effective advertising and marketing content can be created

Blake Irving (GoDaddy CEO) Calls Out Impact of AI on Daily Life in 2016

Blake Irving lists the 4 ways that applied artificial intelligence will make an impact on our everyday lives in 2016. In specific, Irving identifies AI applications in:

  • Vehicles produced by Mercedes and Tesla, that will be used by people en masse on not only highways/freeways but also urban areas. Applied AI in vehicles does not only take the form of autonomous vehicles, it also looks like cars that can assist their drivers in unexpected traffic situations, read speed signs, and avoid hazards on the road.
  • Digital personal assistants like Siri and's Amy. Though we've become familiar with some form of a digital assistant on our smartphones over the past few years, in 2016 we will be able to have more "natural" conversation with our Amys and Siris (e.g. being able to ask "how long will take it for me to get to work if I take the 95" vs. "how long will it take for me to get to work") 

One of the companies mentioned above will actually be present at our 2016 Conference (!!).

Source: LinkedIn

IBM's Watson Applies AI to Diabetes Care

Dosage of insulin injections remains a complex issue for diabetes patients. The interplay between food intake, exercise and insulin leads to many judgement calls about when and how much insulin to inject, followed by more data collection to review the success or issues with such treatment. Now, Novo Nordisk has teamed up with IBM Watson Health to create a "virtual doctor" that will leverage artificial intelligence and draw on a broader set of data.  

Source: Wall Street Journal (paywall)

Bytes on IBM Watson Health: Back in September of this year, IBM opened up its new health services business center right here in Cambridge, Massachusetts. Watson Health was the first targeted effort towards a specific vertical (followed quickly by cognitive computing), but official efforts in insurance, retail and others are sure to follow. The goal? $10 billion in Watson revenue by 2023. 

Gigster Uses AI to Help "Democratize Software Development"

Ever wonder how much it would cost to build that dream app of yours? Gigster can now estimate that cost and give you a production schedule in under 10 minutes. Its artificial intelligence engine gets smarter with each project it completes, which allows the company to efficiently enlist an army of remote developers to plug in pre-made blocks of code. Impressed? So is Andreessen Horowitz, which led a $10 million Series A round just 18 weeks after Gigster launched.

Source: TechCrunch

Bytes on Software Development: Chris Dixon from Andreessen Horowitz wrote about software eating software development over a year ago, claiming that demand will "continue to dramatically outpace the supply." Barriers to software development have already become significantly lower with the proliferation of virtual servers, APIs and open source platforms. Perhaps most disruptive will be the ability for all non-programmers to become software developers, aided by artificial intelligence and services provided by companies like Gigster.