Artificial Intelligence (AI) is the ultimate technology that can be used to design new materials. It can be used to change how we learn and even interact with everything around us. It can change how we manage our world and society, how we create art and how we heal the sick. In short, AI will touch every aspect of human existence. If the last few thousand years have been about development on a human scale, on an industrial scale, then AI will accelerate that scale at an exponential rate that has never been seen before.
That’s because Intelligence is at the heart of all our innovation. It is because of wisdom and creativity that people living in big cities can travel the world and build skyscrapers that reach to the sky. Thanks to wisdom and creativity, people can now read this article on their mobile phones, which is why people can now chat with friends and family anywhere, anytime.
Looking back, the scientific revolution is only 400 years old, and the industrial Revolution only 200 years old. Now, we are in the midst of an information revolution that is only 50 years old. Each of these innovations builds on the last, and it’s changing faster and faster.
The next decade marks the beginning of the intelligent revolution, which is the era of AI. The changes that AI will bring to film, music, television and medicine over the next decade are so incredible that few can fully foresee their future.
AI will become people’s doctor and closest friend; We’ll talk to the AI and ask it in detail where we’re going; When people get sick, we can ask the AI how to treat them.
AI will also reorganize our lives and societies. It will track the spread of disease and optimize the way roads and cities are built. It will also show us insights hidden in the data.
Now, all this has begun quietly.
Artificial intelligence of the future
At this stage, AI is still the product of big tech companies, which spend billions of dollars on research and development. In addition, there are small, super-top teams in the field that provide support in some aspects of data science. But for everyone involved in THE FIELD of AI, it’s still early days.
Large technology companies like Google have armies of programmers who have built a common cloud operating system for Google to run its network of data centers. When developing new technologies, they can easily incorporate AI by updating and adapting their software or inventing new software to manage it. For them, AI is just another cog in the machine.
To get out of the hands of big tech, we need powerful open source platforms that can put AI and ML in the hands of all AI people.
That said, we need AI for Linux.
A host of open source companies, backed by venture capital firms, are working to develop common software to drive AI innovation. Some of them will be giants and others will be swallowed up by the eventual winners as they seek to evolve at different levels of AI’s future infrastructure.
Once we have a clear winner in this area, this stack will be the cornerstone of future AI innovation. When a stack is formed, it allows developers to “move up” to solve more interesting problems. ML will annotate 99% of the data, and humans will examine it. Scientists will start with a series of off-the-shelf methods to solve their problems.
Meanwhile, we’ll continue to see faster and faster chips, driven by the rise of video games and deep learning. These chips will quickly be integrated into the common architecture of every smartphone, game console and sensor.
All of this will drive AI/ML away from the big tech companies and lead to a “Cambrian explosion” of AI applications of all sizes.
These applications, large and small, will lead to rapid growth in agriculture, finance, pharmaceuticals, defense, security, retail, telecommunications and more, and will become even more evident in healthcare.
In the field of healthcare
Ai will revolutionize healthcare in the next decade and beyond, accelerating everything from drug discovery to disease detection and how people get the treatment they need.
Soon, we’ll see AI on the market handling drugs and finding new drugs to fight diseases. Algorithms will design new compounds and new ways to deal with viruses that researchers never thought to try. AI can also detect cancer better than any radiologist, and allow doctors to take their treatments to a whole new level.
The fastest breakthrough will be in disease detection. In the next decade, artificial intelligence will quickly bring radiology to an end.
In 2017, Google researchers demonstrated a 72% accuracy rate for skin cancer detection using the pre-trained Inception V3 convolutional neural network. By 2018, the best classifier achieved 85% accuracy on ISBI open dermatological data sets. By 2020, the best-in-class system will have a 96% accuracy rate, the same as the top radiologists on earth.
Top radiologists have seen AI coming. Writing in Radiology Today, Robert Schier, M.D., describes the Google team’s algorithm that can detect breast cancer better than Today’s best radiologists. He knows very well what this means for his profession: “The emergence of this system marks the beginning of the end of diagnostic radiology.”
The CNN model, which detects metastatic breast cancer from pathological images, has achieved an incredible 99 percent success rate, while human physicians sometimes score just 38 percent on challenging slides.
In 2008, the FOOD and Drug Administration approved an algorithm for medical imaging, and by 2013, that number had fallen to zero. But that jumped to four in 2017, and by 2018, the FDA had approved as many as 18 medical imaging algorithms. It is foreseeable that AI will be great doctors and patients will get better, faster and cheaper care from AI. But as Schier writes, “AI will not ultimately benefit radiology majors.”
In the next decade, we won’t be building star Trek robots, but we’ll be laying the groundwork for machines that know how to quickly detect and deal with everyday health problems.
The most sweeping changes in health care will come from the most unlikely of places. And the COVID-19 pandemic, which began last year, will give biotechnology a big boost.
If the past decade was the rise of big digital technology companies, the next decade will see an unprecedented surge in the power of biotechnology.
The pandemic is an existential threat to people around the world, and even if the disease is not as deadly as we thought, the strain it puts on society as a whole will force healthcare systems to become stronger.
We’ve seen machine learning deployed to track the spread of disease. Data scientists are using GPS data to track how a large bike rally spread COVID-19 to the Midwest. Their model crunches data from weeks ago and reduces the predicted spread time to a few hours, so they can alert governments and regulatory groups in time.
We have also seen the fastest and largest order of magnitude vaccine research in human history. The early clinical results of Pfizer and BioNTech’s vaccines, which are 90% effective, are a remarkable achievement. This is a testament to the drive of open science, fast information sharing and AI for drug design and development.
In the past, it usually took years, even decades, to develop a vaccine. To date, the mumps vaccine holds the record for the fastest development of a vaccine, moving from sample collection to product in about four years.
Emily Waltz writes for the IEEE Spectrum: “As of early September, 34 vaccine candidates have been tested in humans, according to the World Health Organization. Another 145 vaccine candidates are being tested in animals or laboratories. These are quite striking numbers considering that less than a year ago no one had heard of this novel coronavirus.”
Of course, AI cannot speed up the slowest part of drug discovery, the human trial part. But AI is helping scientists analyze viruses and their structures, and showing scientists the details of their attacks.
In the past, scientists spent years studying the structure of viruses to figure out and identify the “breakthrough” that would hit them the hardest. They must thoroughly screen existing drugs to see if they have a chance of killing a new pathogen.
But this time, Chinese scientists were the first to get novel Coronavirus’s genome data and, within a few days, shared it around the world so it could be further studied by people everywhere.
Remember, the human Genome Project took a decade and cost $5 billion, and now a human genome can be sequenced in 48 hours for $200.
The history of the world is sometimes dark and cruel, but also majestic and brilliant. Humans are a highly adaptable and innovative species. We’ve climbed out of the mud, and now we’re sending a spacecraft to Mars, making a vaccine in nine months instead of 10 years, and letting cars drive themselves.
Remember: it took 12,000 years for the agricultural revolution, 400 for the scientific revolution, 200 for the industrial Revolution, and 50 for the information age. The pace of development is accelerating.
Now, as we usher in the dawn of the age of intelligence, who knows where tomorrow’s path really leads? Don’t worry. Where we’re going, we may not need roads.