Credit: Google News
It is less than fifteen years ago when IBM sold its PC business to Lenovo and did the same for its x86 server business back in 2014. It was a major shift for the company that some decades ago was virtually the same as the ‘PC’ itself. But what was the key factor that attracted the company’s attention and made it open a brand new path, different from its past business-safe lanes? Undoubtedly, both businesses that were sold had become less profitable, still, it does not sufficiently explain the reason why a new business had to replace the old one to secure the viability of the corporation.
The narrow victory of the IBM supercomputer, Deep Blue against the Russian chess grandmaster Garry Kasparov in 1997, gave the first signs of the new challenge that the company was ready to seek actively. In 2006, Watson, Deep Blue successor, won two legendary champions on the quiz show ‘Jeopardy’. More recently IBM’s AI took on a human debater arguing for 20 minutes about the need to subsidize preschool – if Cicero were only alive now to see it. Today this is known as the IBM’s Cognitive Business, a more sophisticated way the company uses to underline its determination to be the leader of the Artificial Intelligence era.
The new road is not all roses, as the company faces fierce competition from big techs, like Alphabet, Amazon, Facebook, and Microsoft plus the competition of an enormous startup ecosystem accompanied by huge brain power and strong financial support from private funds. However, with 9.043 patent grants in 2017, compared with 8.090 patents back in 2016, IBM is still on the top of the world list of the most innovative companies with close followers Samsung and LG both from South Korea. Intel is in the fourth place and Alphabet in the sixth with 3.435 and 2.709 patent grants respectively.
Dr. Costas Bekas, Distinguished Research Staff Member, Mgr. Foundations of Cognitive Solutions IBM Research – Zurich, Switzerland points out that 35% to 40% of the company’s R&D budget goes to blue sky research and Project Debater is one example. This is a precious asset for the company’s future as far as many of its competitors prefer to focus on applied research and fast growing business results. But what makes the real difference seems to be the old tight relation of the company with the major business players across all industries. When it comes to new investment, a company likes hearing many different voices, but the voice of a familiar person from the past seems somehow louder and more persuasive.
In this phase of development, AI technology that is actually a machine learning technology, a small only fraction of a generic Intelligence that science fiction writers describe in their stories, needs a lot of training and that means big data from different sources. Dr. Costas Bekas and his colleagues have developed a new Watson service which extracts the most valuable content from the trillions of PDF documents in the world. A U.S. bank piloting the technology reports that what once took 10 days now takes 2 minutes.
This is why it is said that data is the new business capital which is why many companies are very cautious about trusting it safely. So, when Alphabet and IBM knock the same door to ask for data, willing to build a machine learning algorithm to support a company’s operation, maybe an old relation gives the second one a passkey easier.
This is how these days IBM works. It knocks a lot of doors asking companies to use their data to build and improve new algorithms and vice versa: IBM’s clients depend on IBM to help transform their businesses in knowledge-driven enterprises, powered by AI. This is a win-win agreement as the company also benefits from the predictions based on algorithms and IBM learns how to build more accurate algorithms, while the data is securely remaining the property of the client.
It goes without saying that this way of business is not always successful. In October 2013, MD Anderson and IBM announced the development of Oncology Expert Advisor, a clinical decision support technology powered by IBM Watson. However, this project has since been put on hold. Both sides have their arguments but as it proves from later analysis, the divorce came as a result of organizational issues. Health remains one of the most important industries for the company’s business development in the hands of IBM’s Executive Vice President, John E. Kelly III, who wants to channel its intelligence in any business sector.
At the end of the day, through this path of development, IBM gains a lot of experience by using machine learning in the real world. Dr. Costas Bekas further explains that IBM is now in the third generation of Watson development. ‘In the beginning, we knew that this technology had great potential. Then we started to experiment with the system abilities. And now we have several successful business cases, proving that our technology is beneficial for our customers’. One example is Symrise. One of one of the world’s leading manufacturers of fragrances and flavorings is using IBM’s AI to create perfumes based on digital fragrance models. Philyra, as the project is called, has created two fragrances for O Boticario in Brazil and both fragrances will be brought to market by the end of the year.
Moreover, it is expected that a fierce battle will take place in the next years in the field of cloud services. For now, IBM is not ranked first on the cloud business, Amazon and Microsoft have a seat on the two top positions, but it puts a lot of effort to be a leader in this market and the acquisition of Red Hat last year for $34 billion is a clear sign of this effort.
The next big top is how the companies will exploit the most computational power for every dollar they spend on energy. IBM has a multi-axis strategy on this challenge. The blue sky research is always a field full of surprises. Like Alice in the Wonderland, research can bring results that today are considered as pure magic. New hardware is equally thrilling and this is the reason why IBM has invested in the research of quantum computers and the neuromorphic architectures. While IBM is feverishly developing the hardware, algorithmic and mathematical aspects of Quantum Computing, the first commercial-grade neuromorphic processors are expected sooner, according to Dr. Bekas in less than five years. However, quantum computing remains the grand vision, as it has the ability to skyrocket calculations at the scale of googol, which translates to 10 with a tail of 33 zeros.
Credit: Google News