Why is IBM in decline and tries to replicate the human brain?

More than anything else, IBM is famous for reinventing itself. International Business Machines Corp., being established more than 100 years ago, is one of the oldest and most successful IT companies. But the company finds itself in hard times, its stock prices are dropping, core business is weakening, and innovation takes the second place after satisfying shareholders.

In this paper I would like to analyze how IBM got to the current situation and what action is the company taking to tackle its problems.

IBM is and always has been an enterprise services company as it can take care of everything from consulting, software development, hardware development, IT maintenance, to cutting edge research and innovation. Today, there’s much less focus on hardware, but the broad scope remains.

IBM went through a couple of major shifts as the company was evolving in the last 100 years. As the time flew, IBM’s main business was making adding machines, mainframes, PCs, and lately building software and providing IT and business consulting services (after the acquisition of PricewaterhouseCoopers Consulting in 2002 [1]). During these times IBM has showed its ability to accept change and reinvent itself, turning the company into a globally admired brand.

However, in the last years IBM has been marked by cutting costs and reducing workforce. The current situation at the beginning of 2015 is – IBM is in decline. And because current strategies are shaped by past actions, let me briefly dive into IBM’s history.

Why is IBM in decline?

If we go about 30 years back, IBM did a thriving business. As Bridget van Kralingen says [2], the company “had practically invented general purpose computing for business”. It helped put a man on the moon and its research division won several Nobel Prizes. The revenue skyrocketed and by 1984 the company reached the most prosperous times in its history.

Unfortunately, good times were about to end. In 1993 IBM experienced the biggest loss in the history of corporate America – $8 billion [2,3]. The company has missed a number of technology shifts and customers were leaving.

The reason for this big loss was that the business environment has fundamentally changed. “For most of the 20th Century, oligopolies ruled the marketplace. The big successful firms could dictate terms to customers. Firms could proceed on the assumption that whatever high-quality products and services they offered would be bought by the marketplace.” [4] IBM’s hierarchical bureaucracy couldn’t compete with smaller and agile competitors that were truly listening to market needs.

In 1993, IBM was very close to go bankruptcy. Fortunately Lou Gerstner, an external CEO, helped to rescue the company by listening to clients and forming a new strategy. “Gerstner discovered that the very issue that was most bothering the clients (difficulty in integration different technologies) happened to be exactly the kind of service that IBM as a large multi-faceted technology was usually situated to provide… Gerstner saved IBM by a strategic master-stroke i.e. guiding the firm to a competition-free “blue ocean” where its steep, lumbering bureaucracy would not be crippling handicap that it was in its existing businesses.” [4]

As Mark Wilson [3] says, “Most executives are still surprisingly ill informed about what goes on outside the walls of their own business”. Gerstner, coming from outside IBM, was able to see behind IBM’s walls and save the company, but it was only a single one-shot solution. It didn’t cure the underlying disease: bureaucracy. “Gerstner saved IBM but he didn’t institutionalize listening to clients as the driving force of IBM’s culture.” [4]

Another cause of IBM’s current poor performance is its focus on shareholders wealth maximization. In the last few years the company was mainly focused on cutting costs and increasing earnings per share. “From 2004 to 2013, the company spent $92 billion on share repurchases – but only $27 billion on acquisitions.” [5]. This shareholders-oriented strategy was created by the previous CEO Sam Palmisano, who is the creator of IBM’s “Roadmap 2015” strategy, which induced Warren Buffett to invest more than $10 billion in IBM in 2011. [6]

As Steve Denning writes in his article Why IBM Is In Decline [6], Palmisano initially came up with the “Roadmap 2010” plan, which had increased earnings per share from $6 to $10. After accomplishing the plan a new “Roadmap 2015” was introduced to investors with the intention of doubling earnings per share from $10 to $20. In 2012, Ginni Rometty took ever as CEO and continued focus towards the “Roadmap 2015”, which put investors on the first place and caused a lot of problems IBM is facing today.

As industry started to fundamentally change and time when companies needed expensive mainframes and servers was getting over, IBM missed the opportunity to invest aggressively into cloud computing business, even though the company was alert of this shift. Explanation to this lost opportunity can be found in Palmisano words in his HBR interview, “Basically, the shareholders were just asking us to be friendly with capital allocation. They wanted more margin expansion and cash generation than top-line growth, because they knew that if we generated cash, we’d give it back to them in the form of a share buyback or a dividend, not a crazy large acquisition that no one else could see value in.” [6]

In 2012, IBM – known for its great experience with mainframes and government contracting – failed to win the bid for CIA’s cloud computing contract for $600 million. “Its bid was 30 percent lower than Amazon’s, which had no experience with government, yet IBM’s bid was rejected on technical grounds. An appeal by IBM was even more humiliating: it only highlighted how technically inferior its proposal was and so the bid was withdrawn.” [6]

IBM was aware of its lack of technological expertise in cloud computing and started to significantly invest in its cloud portfolio – acquiring cloud services provider SoftLayer for $2 billion in July 2013. Although IBM tried to catch up, Palmisano’s focus on short-term profitability before innovations and long-term investments caused that IBM missed the opportunity to become a leader in cloud computing market. Cloud computing business now makes up only about 3 percent of the company’s revenue. [5]

The decline IBM is experiencing corresponds to challenges that companies face thanks to the age of rapid changes when the window of corporate reinvention is narrow. Palmisano was mistaken when he thought that IT industry is mature and it’s the right time to start giving back to shareholders instead of making long-term investments.

Moon shots for IBM

Regarding Gary Hamel’s Moon Shots for Management [7], here are two challenges IBM needs to tackle in order to reinvent its management and make the company more volatile to the age of rapid changes.

1. Ensure that the work of management serves a higher purpose

“Most companies strive to maximize shareholder wealth – a goal that is inadequate in many respects. As an emotional catalyst, wealth maximization lacks power to fully mobilize human energies… Tomorrow’s management practices must focus on the achievement of socially significant and noble goals”. [7]

IBM’s DNA is focused on moving value toward the margin. All roads must lead to profit. As Panos Mourdoukoutas writes: “IBM leaders have been too busy trying to satisfy shareholders with buyback programs rather than pleasing customers by coming up with new products that beat the competition.”

Shareholders wealth maximization is the biggest weakness of IBM. Since 2000, the company has spent $138 billion on its shares and dividend payments. [8] Contrast this approach with Amazon, when the company’s CEO, Jeff Bezos, told investors 17 years ago that he is going to prioritize long-term market leadership over short-term profitability. The company attracted investors who believed in his long-term time horizon and its stock price has rocketed since that. This year Mark Zuckerberg, Facebook CEO, decided to adopt Amazon’s strategy and introduced his long-term strategy that counts with aggressive investments for the future. Even though this announcement drove loads of investors away and Facebook’s share price fell by 11 percent in after-hours trading, Mark Zuckerberg made sure that his investors were working on the same assumptions. [9]

“Rather than staying on top of the rapid advances in information technology products and services redefining its industry – and every one of its target markets – IBM management had to focus on financial and business reengineering to increase earnings for its investors, as promised… IBM paid a high price for not innovating. It suffered declining revenues for the last 10 quarters – to the point where Rometty had to abandon Roadmap 2015. The superiority of Amazon’s long-term market leadership approach over IBM’s short-term profitability considerations is evident in the two giant’s battle over cloud services.”

The question I’m asking myself, does IBM still stand for its famous “Let’s build a smarter planet”?

2. Eliminate the pathologies of formal hierarchy

“While hierarchy will always be a feature of human organization, there’s a pressing need to limit the fallout from top-down authority structures. Typical problems include overweighting experience at the expense of new thinking, giving followers little or no influence in choosing their leaders, perpetuating disparities in power that can’t be justified by differences in competence… and undermining the self-worth of individuals who have little formal power.” [7]

“To overcome these failings, the traditional organizational pyramid must be replaced by a “natural” hierarchy, where status and influence corresponds to contribution rather than position. Hierarchies need to be dynamic, so that power flows rapidly toward those who are adding value and away from those who aren’t.” [7]

IBM has very steep hierarchy with thirteen layers of management [4]. This causes that the innovator might be ten levels removed from the leader who is supposed to dictate strategy. As Henry Mintzberg explains in Crafting Strategy [10], strategy is nothing that is magically happing way up in the top management, far removed from the details of running an organization on a daily basis. It’s happening out in the field when an employee visits a customer, the product is not quite right, and together they work out some modifications. The employee returns to the company and puts the changes through. After two or three more rounds a new product emerges which eventually opens up a new market. This is how emergent strategy may appear. IBM needs to shift from controlling individuals to enabling self-organized teams where communication works on peer-to-peer basis. This would allow IBM to truly listen to customers, get more value to them sooner and discover new markets.

Current strategy and how IBM wants to reinvent itself

In October 2014, after disappointing quarterly earnings, IBM announced that the company is not going to hit $20 earnings per share in the next year and therefore “Roadmap 2015” has been abandoned. [11] I believe this is great opportunity for IBM to move back into technological engineering and give more flexibility into Rometty’s hands.

Another announcement IBM made recently is that the company is selling its chip-making operations in a deal with Globalfoundries. This is something IBM was always good in, abandoning its businesses that are generating lower profit margins or don’t contribute to the company’s current strategy. As the company did in 2005, when Lenovo acquired its PC business. [12]

As Ginni Rometty mentioned in a shareholder letter [13], IBM strategy has three pillars – cognitive computing, cloud computing, and systems of engagement. She believes this strategy is the right one and that it will help to reinvent IBM. The last ten quarters didn’t look promising mainly because the traditional businesses like hardware, software, and service lines were slowing faster then the new business were growing. “All of which isn’t to say that IBM is doomed. Its security, mobile and analytics businesses are growing. It has a promising partnership with Apple to provide mobile services for companies. It still has plenty of cash to invest and a highly skilled workforce.” [5] The question is when new businesses are going to offset the droppings in traditional businesses. Let’s look more closely into particular pillars of the strategy.

Cloud

Time of mainframes and enterprise servers was replaced with a demand for software as a service available from cloud that offer less expensive ways to store and analyze huge amounts of data. IBM recently released its cloud platform called Bluemix: “Bluemix is an open-standards, cloud-based platform for building, managing, and running apps of all types, such as web, mobile, big data, and smart devices. Capabilities include Java, mobile back-end development, and application monitoring, as well as features from ecosystem partners and open source—all provided as-a-service in the cloud.” [14] Also, Watson – the cognitive technology known for defeating humans at Jeopardy – is now a service and its ability to deal with big data is offered through the cloud. [15] IBM is moving its enterprise software to cloud and tries to attract startup companies and developers with its Watson ecosystem.

Systems of Engagement

This strategic pillar is about enabling people to better use enterprise systems and includes technologies such as social, mobile, and security. IBM recently made a partnership with Apple on business data analytics and with Apple on providing mobile services for companies. Regarding the partnerships with Apple, Brian Klicourse [16] says: “IBM gets a very obvious and really good looking facelift, and can now declare itself to be a leader in the “mobile first” movement in enterprise solutions.”

Cognitive Computing

Cognitive computing makes a new class of problems that are characterized by ambiguity and uncertainty computable. It aims to emulate the human brain’s abilities, action and cognition. For example, human brain is good in understanding natural language and pictures, something what computers still struggle with. Cognitive computing tries to move beyond calculating based on preconfigured rules and programs and get closer to the way in which human brain processes information.

Rometty likes to compare data to a natural resource – “abundant, but worth little unless you refine it.” [15] There are large volumes of data available but companies lack the ability to process and understand them. Mainly because 80 % of this data is unstructured and that’s where cognitive computing comes to play and I believe it’s the most differentiating strategic pillar of IBM’s strategy. IBM has been on the hunt for cognitive system that would understand natural language and could reason since 1997 when Deep Blue, a chess playing computer, won against Garry Kasparov.

Watson, IBM’s cognitive technology that defeated Brad Rutter and Ken Jennings in the Jeopardy in 2011, demonstrated that cognitive systems could learn, reason, and understand natural language. In January 2014 IBM announced that the company is spending $1 billion to launch the Watson Group. Since that time Watson has been used in healthcare or human genome research and more then 3,000 applications to join Watson Ecosystem have been received. [17,18]

Cognitive computing aims to emulate the human brain’s abilities, action and cognition. Watson, although successful at beating humans in Jeopardy, is still recognizably a computer that significantly differs from human brain in several factors like power consumption, size and weight. IBM therefore started a project called SyNAPSE which aims to replicate human brain from the ground up.

Why IBM is trying to replicate a human brain

There are several reasons why IBM wants to replicate a human brain.

First, computers still can’t understand the world as well as we humans can. Since we experience exponential grow of data around us, traditional computers are good in processing and analyzing structured data. “But when it comes to pattern recognition, perception and other seemingly simple tasks, their computing powers pale.” [19] Very simple tasks that our brain can do so well and traditional computers do so awkwardly and not much accurately are recognizing familiar faces in a crowd or driving car in a low visibility.

Second, traditional computers consume a lot of energy. Human brain uses less power than a light bulb when active and even less when resting. Also, human brain is event-driven as it reacts to things in its environment and operates as a massively parallel-distributed processor. Traditional computers do their work sequentially for the most part and are run by a clock. [20]

Third, human brain fills less space than a two-litter bottle and integrates memory and processing together within a fault-tolerant learning system. It also weighs around 1.3 kilogram, which is pretty good if you think about what capacity it offers. [20]

IBM believes that by imitating human brain, we could close the gap between what our human brains are good in and traditional computers struggle with. But to be able to do so, we first need to break the path of traditional von Neumann architecture used for the last 70 years and design a new brain-inspired architecture that operates in a similar way as our human brains.

IBM likes to use the analogy of brain hemispheres. Left hemisphere represents traditional computers and can focus on language and analytical thinking. The right hemisphere deals with senses and pattern recognition. They both make up to a holistic computing intelligence – meaning that brain-inspired computers won’t replace the traditional one but complement them. Today, we have the left hemisphere that is good in crunching big numbers and doing heavy computations. But we still miss the right hemisphere allowing us to sense and analyze patterns. [21]

As ambitious as IBM’s intention can sound, the company and its university partners began working on the project in 2008. The project has received $53 million in funding from the Defense Advanced Research Projects Agency (DARPA). In 2011, first prototype of a brain-inspired chip was developed, following with a programming language and development kit release in 2013. [22]

Dr. Modha, Chief Scientist of Brain-inspired Computing at IBM, says about the project: “This could bring out not just new products but whole new industries. It’s definitely a risky guest. But I can’t think of anything more worthy to do.” [19] This seems like IBM picked up on a trade-up strategy that could create new markets for its cognitive computing business. There are contradictory opinions about whether the project is feasible with our current knowledge of human brain. Michael Jordan, a cognitive scientist and machine learning expert says: “We have no idea how neurons are storing information, how they are computing, what the rules are, what the algorithms are, what the representations are, and the like. So we are not yet in an era in which we can be using an understanding of the brain to guide us in the construction of intelligent systems.” [18] On the other hand, Dr. Modha says: “The brain is proof that this kind of computer can be built. We just need to reverse engineer it.” [19] IBM definitely didn’t pick on an easy quest, but we need to think big in order to make a breakthrough. And brain-inspired computers could change science, IT, and even business in a fundamental way.

How IBM is trying to replicate a human brain

In 2011 IBM introduced the first brain-inspired chip called SyNAPSE. Recently this year, a new generation of the chip called TrueNorth was released. The chip consists of 1 million of programmable neurons, 256 million of programmable synapses and 4,096 neurosynaptic cores. Each neurosynaptic core features 256 neurosn (processor), 256 axons (memory) and 64,000 synapses (communication between neurons and synapses). The chip consumes merely 70 miliwatts and is capable of 46 billion synaptic operations per second. Apart traditional chips which run all of the time, this new neurosynaptic chip is event-driven and operates only when it needs to, resulting in a cooler operation environment and lower energy use. [21,23]

As Knapp explains in his article How IBM’s Cognitive Computer Works [24], programmable neurons (processors) lay out in an array of rows and columns so that the neurons can process in parallel rather than relying on linear structures. The neurons are connected to axons (memory) on the chip by synapses (communication between neurons and axons). The axons act to either excite or hinder the power going through the synapses to the neurons (processors). Depending on the power and information it’s getting from the axons and synapses, the neuron (processor) determines whether it has reached its predetermined “threshold potential” – basically, whether it has found a solution to the problem or part of the problem put to it. If it has, it will “spike” – sending a signal back through the synapse – and reset itself. The synapse then has the solution sent from the neuron, while the neuron goes into a state where it’s awaiting further information. Now picture all the neurons (1 million) acting at the same time, with their signals modulated by the actions of the synapses and axons. All the neurons are working in parallel to each other.

As you can imagine, this architecture is very different from traditional von Neumann architecture where everything is happening based on a linear set of instructions. Therefore IBM created a new special programming language, which is built around “corelets” – an object-oriented abstraction of each neurosynaptic core. In the programming architecture, each corelet has 256 outputs and 256 inputs. Those ourputs and inputs are used to connect the cores to each other. As Dr. Modha says: “Our architecture is like a bunch of LEGO blocks with different features. Each corelet has a different function, then you compose them together.” [23]

“As an example, if you want to use a cognition computer to find a face in a crowd, one corelet might be looking for colors, another might be looking for nose shape. Still another might be looking for cheekbones and so on. Each corelet by itself would run quite slowly, but all of the processing would be parallel.” [23] What could be the other applications of the brain-inspired chips used? For example instruments in the ocean that monitor temperature, pressure, humidity, tides, etc. and determine whether a tsunami warning should be issued. Or search and rescue robots might be equipped with the chips in order to find injured people during emergencies.

Jim Olds, the director of George Mason University’s Krasnow Institute for Advanced Study says that IBM has gotten farther in modeling the brain’s workings than anyone in the world, including his own institute and Henry Markram’s Blue Brain project in Switzerland, which has also received IBM’s support. [19]

IBM’s long term goal is to build a neurosynaptic chip with 10 billion neurons and 100 trillion synapses, but only time will show whether the company will be able to deliver on its plans.

Summary

While investors loved IBM, its “Roadmap 2010” and “Roadmap 2015” took the company too far off course in meeting the needs of its customers. Being too busy with financial engineering and share buybacks, the company missed to invest heavily in cloud computing letting companies like Amazon and Microsoft to lead this area.

The other reason the company is in decline is its hierarchical bureaucracy, which is hopelessly inadequate for the 21st century as its too slow and quite incapable of listening to customers and innovating with them. The role of managers needs to shift from controlling individuals to enabling self-organized teams with peer-to-peer communication. Although parts of the company like Watson Group shows the signs of change, it’s just a small part of the organization.

When Ms. Rometty took over as CEO, she introduced three main pillars of IBM’s strategy: cloud computing, systems of engagement and cognitive computing. The company recently announced some promising movements like partnerships with Apple and Twitter or selling its chip-making operations. Apart from that, IBM is bidding on its cognitive computing as it invested $1 billion in its newly created Watson Group. The company also got into headlines with its brain-inspired chip that could fundamentally change science and IT. IBM’s ambitious plan is to close the gap between what human brain is good in and computers struggle with. The goal is to create “neurosynaptic supercomputer” which is using both traditional and brain-inspired chips enabling it both to crunch numbers and quickly analyze real-time patterns.

We will see as the 21st century unfolds whether brain-inspired technology has the potential to change the paradigm and open up new industries. But there is one thing we can say for sure – if IBM wants to survive another 100 years, it will have to make radical changes to its management practices so that it’s able to adapt on a continuous basis, not just in times of major crisis.

Bibliography

[1] IBM. IBM, PricewaterhouseCoopers Complete Sale of PwC Consulting. [Online]. https://www-03.ibm.com/press/us/en/pressrelease/491.wss
[2] Bridget van Kralingen. (2010, July) IBM’s Transformation – From Survival To Success. [Online]. http://www.forbes.com/2010/07/07/ibm-transformation-lessons-leadership-managing-change.html
[3] McKinsey & Company, “What strategists need: A meeting of the minds,” McKinsey Quarterly, September 2014.
[4] Steve Denning. (2011, Nov.) Will IBM Survive Another 100 Years? [Online]. http://www.forbes.com/sites/stevedenning/2011/07/11/will-ibm-survive-another-100-years/
[5] David Shipley. (2014, Oct.) Does Watson Know How to Save IBM? [Online]. http://www.bloombergview.com/articles/2014-10-22/does-watson-know-how-to-save-ibm
[6] Steve Denning. (2014, May) Why IBM Is In Decline. [Online]. http://www.forbes.com/sites/stevedenning/2014/05/30/why-ibm-is-in-decline/
[7] Gary Hamel, “Moon Shots for Management,” HBR, February 2009.
[8] Panos Mourdoukoutas. (2014, Oct.) IBM, McDonald’s And Coca-Cola: Right Companies, Wrong Strategies. [Online]. http://www.forbes.com/sites/panosmourdoukoutas/2014/10/25/ibm-mcdonalds-and-coca-cola-right-companies-wrong-strategies/
[9] Chunka Mui. (2014, Nov.) Zuckerberg Adopts Bezos’ Strategy To Avoid IBM’s Fate. [Online]. http://www.forbes.com/sites/chunkamui/2014/11/11/zuckerberg-adopts-bezos-strategy-to-avoid-ibms-fate/
[10] Henry Mintzberg, “Crafting Strategy,” Harvard Business Review, 1987.
[11] Julie Bort. (2014, Oct.) IBM’s Ginni Rometty Just Confessed To A Huge Failure — And It Might Be The Best Thing For The Company Read more: http://www.businessinsider.com/ibm-gives-up-on-its-2015-promise-2014-10#ixzz3MNYXlg74. [Online]. http://www.businessinsider.com/ibm-gives-up-on-its-2015-promise-2014-10
[12] Steve Lohr. (2014, Oct.) Weak Results at IBM as Its Strategy Shifts. [Online]. http://www.nytimes.com/2014/10/21/business/amid-a-shift-in-strategy-ibm-reports-weak-earnings.html?_r=0
[13] Larry Dignan. (2014, Mar.) IBM CEO Rometty talks reinvention, strategy. [Online]. http://www.zdnet.com/article/ibm-ceo-rometty-talks-reinvention-strategy/
[14] IBM. IBM Bluemix. [Online]. http://www-01.ibm.com/software/ebusiness/jstart/bluemix/
[15] Jonathan Berman. (2014, Nov.) IBM’s Emerging Market Strategy Has 3 Pillars. [Online]. https://hbr.org/2014/11/ibms-emerging-market-strategy-has-3-pillars
[16] Paula Rosenblum. (2014, July) The IBM Partnership Makes Apple A Real Enterprise Player In Retail. [Online]. http://www.forbes.com/sites/paularosenblum/2014/07/18/the-ibm-partnership-makes-apple-a-real-enterprise-player-in-retail/
[17] Brad Power. (2014, Aug.) How Watson Changed IBM. [Online]. https://hbr.org/2014/08/how-watson-changed-ibm/
[18] Gil Press. (2014, Oct.) IBM Reinvents Itself By Establishing A Frontier Homestead For Cognitive Computing. [Online]. http://www.forbes.com/sites/gilpress/2014/10/27/ibm-reinvents-itself-by-establishing-a-frontier-homestead-for-cognitive-computing/3/
[19] Andy Greenberg. (2009, Nov.) IBM’s Cat-Brain Breakthrough. [Online]. http://www.forbes.com/2009/11/18/ibm-brain-science-technology-breakthroughs-supercomputer.html
[20] Roger Kay. (2011, Sep.) Cognitive Computing: When Computers Become Brains. [Online]. http://www.forbes.com/sites/rogerkay/2011/12/09/cognitive-computing-when-computers-become-brains/
[21] IBM. IBM Brain Power: The origins of a brain-inspired chip. [Online]. http://www.research.ibm.com/cognitive-computing/brainpower/
[22] Alex Knapp. (2014, Aug.) IBM Builds A Scalable Computer Chip Inspired By The Human Brain. [Online]. http://www.forbes.com/sites/alexknapp/2014/08/07/ibm-builds-a-scalable-computer-chip-inspired-by-the-brain/
[23] Alex Knapp. (2013, Aug.) IBM Develops Programming Language Inspired By The Human Brain. [Online]. http://www.forbes.com/sites/alexknapp/2013/08/08/ibm-develops-programming-language-inspired-by-the-human-brain/
[24] Alex Knapp. (2011, Aug.) How IBM’s Cognitive Computer Works. [Online]. http://www.forbes.com/sites/alexknapp/2011/08/26/how-ibms-cognitive-computer-works/2/