WHERE DO WE STAND TODAY?
Today, anywhere you look, startups are emerging. The numbers are simply mindboggling! Extrapolating from the data that Dr. Paul D. Reynolds, Director of Research Institute, Global Entrepreneurship Center, provided, we find that there are:
472 million entrepreneurs worldwide attempting to start 305 million companies, approximately 100 million new businesses will open each year around the world.
How crazy is that! 100 million new businesses each year, that is, over 273972 new businesses per day! Further, out of the 100 million, 1.35 million are tech startups.
But that isn’t the worst of it – 9 out of 10 startups fail!
Why this staggeringly high failure rate? CB Insights performed a study to find the top 20 reasons for failure.
42% of the failures are due to “No market need”. People make products that consumers aren’t willing to buy. They fail to study the market needs properly before plunging into their project, thus flushing down millions of dollars.
The same mistake cannot be repeated with the creation of Artificial Intelligence. Hence, we have to first study what the world needs and accordingly develop intelligent systems. We need a vision of the future of AI before we plunge into its creation.
A VISION OF THE FUTURE OF AI
Science and technology have changed our lives with staggering effect. In around 100 years, the average life expectancy of a human being has increased by nearly 40 years.
And yet, it is not enough. The following data from WHO reveals that the leading causes of death in the world are: heart disease, stroke, chronic obstructive lung disease and lower respiratory infections.
What if we could predict these diseases and thus prevent them? Or more importantly, how do we predict?
And this is where AI steps in.
Researchers today are developing machine learning techniques which analyse huge amount clinical records to predict imminent diseases. These programs sift through the medical history of thousands of patients with a particular disease, then look for others with similar records, and gives the likelihood of the disease occurring. This paper is an example of AI being used to predict heart disease.
Other researchers are combining machine learning methods with advanced MRI techniques to help predict Alzheimer’s disease, brain cancer and other diseases of the brain. The machines learn to recognise patterns in the scans and extrapolate to predict the diagnosis. Today, researchers are able to predict the Alzheimer’s disease with 82-90 percent accuracy.
Some are even studying genes, attempting to get the relation between a particular gene and a disease. As the other techniques, they too are applying machine learning techniques to crunch tons of data to extract patterns that might help them find the root of the disease.
Surgical robotics is steadily emerging too. Today, there are hospitals where basic surgeries are performed without a doctor slicing into the patient directly. An operator guides robotic arms with the help of joysticks. These arms reduce jerky or shaky movements, thus increasing precision. With the addition of superior visualisation, surgical robotics minimises incisions, thus reducing risk and need for medication. The da Vinci System has performed over 3 million minimally invasive surgeries successfully.
Furthermore, smart bionic limbs are using machine intelligence to aid invalids lead a normal life. They sense and adapt to the environment and predict the user’s intentions to provide greater stability and ease.
AI is not only helping in diagnostics and surgeries, but also in designing drugs. Atomwise’s AtomNet studies protein structures, which can be considered to be “locks”, and tries millions of molecular combinations to open these “locks”. Basically, it’s designing complex molecules to destroy harmful protein combinations, or put differently, designing drugs to cure diseases.
Though these technologies are still premature, we can see the potential they have. In the future, life expectancy is bound to rise, diagnostics will be more accurate and easily obtained, surgeries will be automated, medical advise and care will be provided by virtual assistants online, drugs will be more effective with minimal side effects, smart exoskeletons will aid the disabled to lead a normal life. We may now even dream of exoskeletons merged with the human body for superior performance, human augmentation, automated gene modification, advanced cyborg technology, nanobots racing through our blood streams clearing clots and cleansing the body of diseases and other crazy ideas. Man, how cool will our future be!!!
What if you had a friend for life, a friend you would never lose, a friend who understood every emotion, every idea, every intention? What if this friend was unique to you, would guide you, be there to support and help you? What if I told you that this is our future?
With the advent of Internet, there is a boom of information, so much information that discerning the useful ones from the worthless ones is becoming nearly impossible. What if somebody could do extract the meaningful content and use it to make our lives easier?
This should be the aim of virtual personal assistants (VPA). Today, we have Siri, Cortana, Google Now, Watson – all crude versions of our vision. The technology currently is basic, involving pattern recognition, knowledge bases, natural language processing and sentiment analysis among the numerous other techniques. And yet, true “cognisance” is yet to be achieved.
In the future, here are some tasks we might expect our VPAs to perform: schedule meetings and manage time, monitor personal health and in case of emergency automatically call medical aid, connect home, car and office through IoT, update user on news, traffic and weather, book reservations or order from restaurants, provide information requested by the user by searching the Internet, but more importantly, customising and adapting to the user’s needs. There are tons of other tasks, but listing them all would be mindless.
But we want our VPAs to be more than that. We need them to understand our emotions, our sentiments, our moods…when we are down, automatically soft music should play and the lighting of the room should mellow down, where we are jovial, upbeat music should cheer us on etc. Furthermore, our VPAs should be able to converse meaningfully, understand our feelings and guide us by extracting information from the Internet. Stop for a moment and ponder, what is it to really understand? How to make a machine understand our feelings?
This is merely a glimpse of the future of virtual personal assistants, but one thing can be sure: they will have a significantly large role in our future. Currently, I’m creating my own virtual personal assistant in Python and will soon be writing about it.
Cyborgs, Humanoids and Robots
Some readers now might be “Ah, now he’s talking about AI”. This is the typical vision of AI, for it has some truth. The personal assistants need not only be virtual, but will soon have bodies. They will aid us in our daily mundane tasks, such as driving the kids to school, throwing out the trash, playing, helping cook, helping wash and any other task you could think of. Furthermore, they could help the aged with daily activities. Today, humanoid robots are already emerging, such as Asimo, Nao, Atlas and Actroid-SIT.
Furthermore, there is a tremendous boom in industrial robots which automate manufacturing. Today they are used for welding, painting, picking and placing, packaging and numerous other tasks. They are in demand because of their precision, speed and endurance.
Apart from these, robotics will soon have applications in military, space exploration, agriculture, medicine, sports, fire fighting, construction and in innumerable other fields. Lastly, we should keep an eye out for nano robots too, for they are an emerging technology.
By 2018, autonomous cars will be rolling on the streets created by companies like Google, Uber and Tesla. These automobiles use machine vision, GPS and odemetry among other techniques. There is a lot of research in this field because, as shown by the graph above, deaths due to road injuries ranks 9th, with 1.3 million dying per year. Reports say that driverless cars could reduce road fatalities by 90%. Furthermore, in the foreseeable future people will avoid buying cars because automobile services like Uber will be a tap away and considerably, and this will cause a reduction and a better organisation of traffic, which could further reduce fatalities.
Other autonomous vehicles too will soon emerge, like Hyperloop, drones, hovercrafts, trains, planes, ships etc.
AI will cause a boost in business. With the help of predictive analysis, there will be major improvement in stock market prediction, business models, recommender systems and numerous other fields.
Today, Amazon, Facebook, Microsoft and Google, among numerous others, are using AI to analyse consumer behaviour to provide better ads, services and products.
These are merely some of the fields in which AI is going to have a major impact. This graph reveals the future of emerging technologies:
To conclude, AI is currently undergoing a boom and therefore, an AI startup will have high demand and relatively easy funding. And thus, guided by our vision, let us proceed to create our future.