4-9 Surviving in the New Information Economy

It never occurred to me that I would someday become omniscient, which I have for all practical purposes. Information flows through the Internet at the speed of light. We are living through an explosion of information in its many forms. The economy is going digital, and programming skills are in great demand at many companies. As the world shifts from an industrial economy to an information economy, education and job training will have to adapt. We are witnessing the beginning of a major shift in how we live. The Age of Information will have as profound an impact on society by enhancing our cognitive abilities, as the Industrial Revolution had 250 years ago by enhancing physical power. The timetable for this transformation will be much more rapid than the Industrial Revolution, which took 100 years for the world to adapt to machines. The Information Age will play out over the next few decades. We heard last week that the South Korean Go champion was beaten by AlphaGo, an artificial intelligence program based on deep learning that was bio-inspired. Learning in deep neural network models has become possible because of the million times increase in computer power since the 1980s. Deep learning has made it possible to reach human levels of performance in speech recognition and object recognition in images. As we reverse engineer the brain’s learning systems, we are also gaining new insights into human intelligence. Nature evolved learning as a way for a broad range of problems to be solved by experience with the world, and humans learn faster and more deeply than any other species. Google recently released a new version of Google Translate based on deep learning, which was a quantum leap in the quality of translation between natural languages. Literally overnight, language translation went from a fragmented, hit-or-miss jumble of phrases to seamless sentences. Previous methods searched for combinations of words that could be translated together, but deep learning looks for dependencies across whole sentences. When you use an Android cell phone to transcribe speech, you are using deep learning. When speech recognition is combined with language translation, it will become possible to communicate across cultures. Star Trek’s universal translator is within our reach. An even larger disruption is beginning to affect service industries and professions, as machine learning matures and is applied to many other problems where big data are available. Medical diagnosis based on the records of millions of patients will become more accurate. A recent study applied deep learning to a medical database of dermatological images for over 2000 different skin diseases. The network was trained to diagnose these diseases. Performance was comparable to that of expert dermatologists. It should now be possible for anyone with a smartphone to take a photo of a suspicious skin lesion and have it diagnosed instantly in the cloud, a process that now requires a visit to a doctor’s office, a long wait for the lesion to be screened by an expert, followed by a substantial bill. This will greatly expand the scope and quality of dermatological care. Deep learning is just beginning to affect the legal profession. Much of the routine work in law offices that charge hundreds of dollars an hour will become automated. Complying with the increasing regulatory complexity will also benefit from automated systems. Legal support will be available for the average person who cannot now afford a lawyer. Not only will legal work be cheaper, it will be much faster, a factor that is often more important than the expense. Deep learning will have an even bigger impact on your everyday life. As you interact with the Internet, you are generating data about yourself that is machine-readable. This can be used to create a personal assistant that knows you better than anyone else in the world, and will not forget anything. Your children will have their own personal tutors who’ll accompany them throughout their education. The educational opportunities for the children of today’s children will be better than the best education available today to wealthy families. As these new technologies mature, the new jobs that will be opening up will require new skills. The steam engine displaced human labor, but created jobs for engineers who could build and service steam engines. As jobs that now require cognitive skills are taken over by automated AI systems, there will be new jobs for those that can create and maintain these systems. These new jobs will require new skills. So, be prepared for a lifetime of learning.