Needless to say, machines, or the ideas of them, have fascinated us for all of known human history. And today we have machines that can at least imitate thoughts if not have original ones. Without knowing it AI (or Artificial Intelligence) touches everyone’s lives in some way, pretty much on a daily basis. The ads you see, the manipulation of big data, the fuzzy logic sensor that washes your clothes… almost every service, smartphone and mass produced product has varying levels of AI involved in the process, somewhere.
With AI taking over every field, it’s no surprise to anyone that startups working on AI are hot right now. The future is AI, and the future giant companies of the world (the Facebook, Apple and Google of tomorrow) might start off as a small AI startup today. That said, Google, Microsoft, IBM, Facebook, Apple, Uber, and even Shell are investing heavily in AI.
But enough of that. Let’s take a look at the most interesting AI startups that we feel are doing very interesting work.
Echobox is helping machines understand the meaning of words
Founded by two Cambridge graduates, this startup claims to have developed the world’s first AI that truly understands the “meaning of content”. Echobox understands the constantly evolving landscape of social media so as to try and maximise every brand’s reach. Even casual users of Facebook and Twitter know that there are a number of different variables that allow one’s post to have the maximum possible reach. Echobox takes into account all of these variables so as to direct different content to the required target audience. It is a sign of the times that social media accounts have gone from being run by adolescent interns to being a campaign orchestrated by an AI. Clients such as Le Monde, Le Figaro, Axel Springer, and San Jose Mercury News are all seeing the benefits of using Echobox.
Viv Labs is trying to redefine how we talk to our gadgets
In 2016, Samsung agreed to acquire Viv – an assistant system created by the same people who made Siri. When one reads a sentence like that, the first question that comes to mind is,”How is it different from Siri?” Well, the difference is twofold: First, it allows for more complex, conversational queries because it assimilates the information gathered by all the different applications on the phone to interlink into one vast web of information it can work on. Secondly, because it can literally write its own code, Viv can create and apply new programs without ever having heard of them before. Its founders say that Viv is a “software that builds itself”.
Millions of humans all over the world go to spiritual retreats to achieve ‘self-actualization’. This sort of sounds the same doesn’t it? A magnificently scary thought.
Mad Street Den – Redefining fashion and lifestyle industry
Launched in 2013 by wife-husband duo Ashwini and Anand Chandrasekaran, Mad Street Den was originally based out of Chennai. Within its ranks, Mad Street Den has a fascinating group of Data Scientists, Neurologists, Designers, and all round computer geniuses. They aim to build a visual identification intelligence so that the fashion and lifestyle industry is able to better customer relations. They want to also build tools to accurately understand and predict the way a customer behaves, thereby giving the corporation a competitive advantage.
An Indian origin startup with offices all over the world, Mad Street Den are killing it in the AI world
This is a relatively newer startup that acquired their seed funding in 2016. However, helmed by a CMU graduate, it has to potential to become a global entity. Already it has a physical representation in Europe, the US, and India.
Banjo is creating a real-time map of the world
This startup has a really bold aim – to create an Alexandrian Library worth of information via social media. It was founded in 2013 after the Boston Marathon bombings, and sounds like the real world manifestation of the AI that controls everything in Jonathan Nolan’s Person of Interest. Creating a web of information that includes Facebook, Twitter, Russia’s VKontakte, and China’s Weibo, it can classify locations and photos to quite literally ‘rewind’ the flow of information to accurately plot past events.
Take, for example, a terrorist attack. Banjo could be used to create a blueprint of the situation using the millions of social media updates available to it. It would essentially be able to play with our fourth dimension whenever it pleases.
Defined Crowd is building bridges between mankind and AI
Siri, Cortana, and the aforementioned Viv all need millions of hours of well-curated language input. Defined Crowd is filling this niche need by creating complicated pipelines that accurately filter and interpret data. This is done partly through their AI, and partly through actual real human beings. Amy Du, Co-founder and CEO, said that there is a very subtle difference between an excited tweet and a happy tweet. This is where the company’s crowdsourcing ability comes in. They’re already in almost 50 countries with 50-100 people working on language data in each of these countries. After the initial kinks are ironed out from each country and their human workforce start delivering consistently, they receive almost real time information about the language.
It is part of the Microsoft Accelerator program, which means they have a bright future. Besides, it’s a heart-warming (CPU-warming?) story about robots and humans working hand-in-hand? We’re going to need a few of those to convince the superintelligence that we’re useful!
Stepsize Layer helps developers make better judgement calls
Stepsize Layer is a desktop application that automatically adds context to code by using its tools to develop software and attach historical data to a piece of code. Using this AI tool, Developers can select a piece of code, run it through Step Size, contact the original author and find the context. This will ultimately lead to a community of coders where all code is properly contextualized and connected to all other pieces of code. Step Size is fully integrated with Slack so that it can use it to do its job.
Galaxy.ai : Your personal health data avatar
“When symptoms appear, it is already too late.” Galaxy.ai uses its own proprietary software to aid the world to become a better place to live in. It collects data from traditionally difficult areas such as the sectors of medical care and insurance to help doctors worldwide make better choices. They are building the world’s first ever AI prognosis engine that will help doctors treat patients better. It also helps provide better, more affordable insurance by studying the data provided by insurance companies.
Founded in 2015 and based out of New York, they are building the proverbial Health Bot – something that is a common occurrence in loads of video games.
BenevolentAI accelerates scientific discovery
Every 30 seconds, a scientific paper is published. Over 10,000 a day are added to just the medical portion of research. Because of this ungodly number, the amount of research that becomes ‘usable’ is only a small portion of this. BenevolentAI wants to solve just this by developing a tool that combs through all this data to provide researchers and scientists with up-to-date information. This leads to a much faster rate of drug discovery, which in turn leads to a better healthcare system for all.
While Galaxy.ai and BenevolentAI don’t work together, they are theoretically the stuff sporting dream teams are made of. One enables medicine makers while the other diagnoses diseases. A one-two punch that no virus can evade. Here’s to universal healthcare!
Niki.ai – An Indian AI startup funded by Ratan Tata
This Indian firm was founded in 2015 by 4 IIT Kharagpur graduates and is a chatbot that allows for a simpler customer experience. Its core functionalities include mobile recharge, travel services, and all bill payments. Ratan Tata recently gave Niki.ai a hefty amount of seed money, so they must be doing something remarkable. While this doesn’t sound unique, their natural language processing power is aimed specifically at how Indians talk and type. It has partnered up with Ola, Uber, as well as PayTM to provide their facilities through their app. Currently, it provides services to around a hundred thousand customers.
One of the few Indian AI startup, they procured their seed money from none other than Ratan Tata
Clarifai is an image & video recognition API
You know how Facebook either asks you to or automatically tags people in your pictures? Remember how often it’s wrong? Clarifai is hoping to have a foolproof solution for that. It claims that it can teach its AI to identify an object with less than 10 input images, without additional code to help it. It also keeps collecting data and learns as it goes.
|Back in 2016, Google Photos labelled an African-American couple as ‘gorillas’. If only Google was aware of Clarifai at the time. They could’ve avoided something like that altogether.|
What we’ve presented here is merely the tip of the iceberg. There are probably hundreds more who we haven’t even heard of yet but may potentially hold the key to finally achieving general AI – an artificial super intelligence that is as intelligent as a human (or more). Whether we like it or not, 2017 is the year when thousands of companies will work on AI more seriously than ever before. Humans don’t come much smarter than Stephen Hawking, and he’s already stated, “The development of full artificial intelligence could spell the end of the human race.”
Of course, Hawking is only human and fallible, so we can only hope he is wrong this time because none of the companies we’ve listed here or thousands more that we haven’t been paying any attention to his warning… Should they be? What’s your take? Write in and tell us before these bots take over our jobs!
This article was first published in April 2017 issue of Digit magazine. To read Digit’s articles first, subscribe here or download the Digit app for Android and iOS. You could also buy Digit’s previous issues here.