An AI Playground

This week I was invited to the official opening ceremony of the ARIC (Artificial Intelligence Center) Hamburg. The ARIC brings together companies, start-ups, research institutes, banks and politics to initiate AI-based projects and establish AI solutions on the market. Besides good conversations, I experienced interesting presentations introducing AI projects.

A very large established finance company uses AI in two ways. There are short-term (in 1 to 3 years duration) projects in which modern applications and new user interfaces are developed. In the long term, in cooperation with ARIC, completely new business areas are tackled and the old processes are fundamentally improved, e.g. in the analysis of legal documents.

A communications company presented how they use AI to evaluate and optimize the efficiency and reach of marketing methods. A consulting firm showed how AI in image analysis can be used to categorize defects in aircraft engines much faster.

There are many ideas on how AI can drive new business, and yet it seemed a bit like a playground to me. This is not meant negatively. It’s about playful experimentation. There will be many more experiments to try. And it’s about starting on a small scale and proving the value of AI solutions, as I wrote earlier.

The more AI-based business models work, the more new ideas are coming up. I can imagine that AI will become much more interesting for many companies. And faster than you might think.

Truly Intelligent Machines

The definition of artificial intelligence can be vague. Sometimes it seems to be just brute force number crunching. There, more and more computing power is used to create a behavior that seems to show intelligence. But if we look behind the scenes of Deep Blue and other supercomputers that master games like chess or go, these are special cases where knowledge is optimized in a clearly defined area.

Human intelligence is much more creative and adaptable. It is prepared for every eventuality in our lives, much more than any computer.

And this is exactly where the 15-year-old classic by Jeff Hawkins and Sandra Blakeslee comes in: “On Intelligence” is a book in which we learn in great detail how the human brain works, how the neocortex is structured, how we use it to remember things, and how we make decisions. And it is precisely this biological template that the authors use to give us clues as to how to build truly intelligent computers.

A colleague and friend recommended this book to me, and I can only pass on this recommendation. Even if the predictions of 15 years ago did not really come true, it is still an enlightening reading.

“The most powerful things are simple,” Jeff writes in the prologue. He’s right, you might just think of the iPhone. So this book presents a simple and straightforward theory of intelligence. It is very profound when the individual cells and cell regions in the brain are explained how they interact and how information is stored and retrieved. Yes, you should concentrate while reading, but is it also understandable for non-neuro-scientists.

Now, if a machine uses this behaviour of the human brain, then it is really intelligent. Jeff assumes in this book that in 10 years (that would be 2015) such intelligent machines will exist. But in the next sentence he gets more cautious because it might take longer.

Jeff calls for the construction of such machines, which have the human neocortex of the brain as a model. In the book there are some examples, e.g. how such machines communicate and capture the world’s weather in a level of detail that seems impossible today. Do we really want that? I’m not sure that’s a good idea. And I haven’t heard anything more about such machines.

Anyway, I recommend the book “On Intelligence” to anyone interested in intelligent computers. You’ll have more respect for your brain after reading it.

Investing in AI and The Role of VMware

At the NORTEC 2020 trade fair for Manufacturers I was invited to a round table discussion about the introduction of AI in the manufacturing industry. Large companies, universities and local business leaders explored how to use AI to drive innovation and create a business plan for it. I was asked to present the role of virtualization for AI/ML projects, and a data scientist was interested (and surprised) by the performance benefits of virtualization as described in the VROOM blog article How Does Project Pacific Deliver 8% Better Performance Than Bare Metal?

Several representatives and local executives from private and family businesses discussed their business. Small and private companies are driving the economy in Northern Germany, where there is not a single DAX company, but many small and medium-sized companies. I was surprised to learn that these smaller companies increase their turnover much faster and more strongly than the large public companies. The consensus was that long-term investments exceed short-term investments. Public companies must take shareholder value into account and provide quarterly figures. Many decisions are made to increase short-term revenues. Smaller companies have a time horizon of 10 to 20 years for their investments, resulting in a more stable and reliable business. They work over many generations.

This has an interesting influence on their AI strategy. These companies cannot afford large investments, so we have discussed joint projects with students from local universities. These entrepreneurs cannot risk investing large sums of money because they have to control the risks. But they are very interested in AI and there are first companies that are starting to get value out of AI. But they are only at the beginning. Another stumbling block is the concerns about using the public cloud for AI projects, especially in terms of compliance and intellectual property protection. As a result, they will want to run AI/ML software in their local data centers or locations. The amount of investment is often only around €10,000 to €15,000 for hardware, so at first I thought this was too uninteresting for cloud infrastructure providers like VMware, who tend to support larger projects. But I was asked about the virtualization of AI/ML workloads because almost everyone has had good experiences with VMware vSphere (or VMware Workstation). In addition, universities and research institutions like DESY have to cover completely different dimensions, which can make infrastructure projects with virtualization interesting.

Unexpected Side Effects

In the podcast Die Maschine: Kontrolle ist gut, KI ist besser (in German language, by the radio station Deutschlandfunk) a scary fictional story is told from the 21st minute on:

An artificial intelligence has been developed that controls and executes all drug shipments worldwide. Because this was so critical, a special algorithm was chosen to ensure that individual population groups are not disadvantaged under guarantee, an algorithm which is always 100% politically correct.

When the artificial intelligence was activated, things went well at first, but then the number of deaths of diabetics increased in the rich countries. Insulin is lacking everywhere in the hospitals of the industrialised countries. How could this happen?

Well, the system worked exactly as it was designed. However, the artificial intelligence took into account the need for drugs worldwide. But there were not enough drugs like insulin for everyone on earth. Underserved areas, especially in Asia and Africa, received more drugs from the artificial intelligence, while the rich countries received less. Thus the shortage was distributed evenly across the globe.

It is a similar dilemma to two burning houses with people trapped inside, but you only have enough helpers to fight one fire and save the people, not both fires. What are you doing?

These are ethical questions that an artificial intelligence cannot answer automatically. So when artificial intelligence is used to sustain life, we should look very closely. And well-intentioned is certainly not always well done, as the story of the drugs shows.

Between Hype And Extraordinary Benefits

Last week I was at the ML Conference in Berlin. The conference was very instructive for me. The lectures were about the technology behind machine learning, data science, security and algorithms. But then it was also discussed how to set up an ML project, which mistakes can be made, and why container technologies like Docker or Kubernetes make ML projects much easier. Most exciting were the lectures about concrete applications of machine learning and artificial intelligence. And of course the discussions with other participants did not come too short.

My Conclusion

Most companies are currently experimenting with machine learning. There are only a few ideas on how these new technologies can bring real business benefits today. The hype in the media – both positive and negative – was far away. At the same time, everyone I spoke to agreed that in a few years, any company will have to use machine learning to stay in the market. It will be the lowest common denominator.

It will be used both in the public cloud and in the company’s own data center. The biggest problem in your own data center seems to be the slow processes. “It takes far too long for me to get a system set up. I’m much faster in the public cloud,” several people told me.

From VMware’s (my employer’s) point of view, I noticed that some participants knew us, but none associated VMware with machine learning and artificial intelligence. VMware offers interesting options when it comes to the flexibility and security of ML applications. Well, I’ll deal with that another time.

Here are two more highlights from my point of view:

How we avoid maritime accidents

The Keynote Data to the Rescue! Preventing Accidents at Sea by Dr. Yonit Hoffman of the Israeli startup Windward inspired me. Did you know that 10% of all ships have an accident once a year? With big data analysis and machine learning, Windward has developed models to predict and prevent accidents. 150 million data points on ship positions, ship types, weather information, water depths, proximity to ports, and many more have been brought together and translated into deep knowledge and understanding of what happens at sea. The lecture was really captivating.

The results have led to a very fruitful and interesting cooperation with insurance companies. Imagine which worlds meet. Maritime insurance companies have been working in a similar way for 350 years – often with paper. Now something new is emerging. And it is clearly driven by business. This is how I imagine the future of machine learning.

Rainwater drainage pipes for China

The most exciting conversation was with an employee of a company that manufactures sanitary systems for large buildings and rainwater drainage pipes. A special challenge are large flat roofs, e.g. for industrial plants or shops. The drainage pipes must have exactly the right diameter. If they are too small, too little water flows off. If they are too large, there is no suction effect. With the optimum diameter of the pipes, a suction is created in the pipe, which optimally drains off the rainwater.

I have learned that many years of professional experience are necessary to design these optimal pipes. This is a problem when you want to enter new markets. Now the company has managed to determine optimum rainwater pipe diameters much more quickly using machine learning. Now, for the first time, it is possible to enter new markets such as China without having employees there who have decades of experience.

If machine learning and artificial intelligence already deliver value in this extraordinary way today, then we will certainly see even more examples in the future. The limits are our creativity and imagination. Knowledge of technology is not enough. This is only a necessary, but not a sufficient condition for the successful use of machine learning. Identifying business ideas is the real challenge.

DeepL – AI Translates Best

I write my articles first in German and used Google Translate to create the English version. That was about 80% good, a bit polished, then the second language was online.

A colleague told me yesterday that DeepL Translator could be a much better translation engine for me. I tried it, and I’m thrilled. In a direct comparison with Google Translate, DeepL is much better at translating my German articles into English.

What makes DeepL so special? I think it’s the consistent use of neural networks, machine learning and artificial intelligence. This is described on the DeepL Press website:

“2016 – The team begins to construct a neural network capable of translating any kind of text. It later becomes known as DeepL Translator. This system builds on the team’s years of experience with machine learning and natural language processing, and has been trained with over a billion high quality translations collected by Linguee’s web crawler.”

Now another 3 years have passed and DeepL is getting better and better. That’s what I love about good technology. Keep it up, DeepL Team!

The Race Between Regulation And Innovation

While politics and the economy are still discussing how Artificial Intelligence (AI) should be regulated, the development of new AI solutions is progressing. It’s a race. New and better robots are being built, for the kitchen and for the military.

Algorithms take over more and more decisions. Certain groups of people may be disadvantaged, as happened in the US in October 2019. The algorithms are not “evil” per se. They reinforce the prejudices that people already have. They open relentlessly what we think, because our thinking influences the data sets and the AI ​​rules (mostly unconsciously).

I believe that new regulations will only help to a limited extent. We already have very good laws that make sure that all people are treated equal and no harm is done. The application of existing laws and regulations is important. And more importantly, to become aware of how AI systems are learning and how we can ensure that AI systems comply with current regulations.

I suspect there is still a lot of research to be done. To steer the behaviour of AI systems in the right direction is an opportunity for us to shape the future positively.

Does that scare you? Well, what would be the alternative? To ban the use of AI systems? I think that would be worse, because it could create a shadow economy that would develop AI systems further. Just as there are drug cartels that do illicit business. But with AI it is much more dangerous if it is developed in secret, perhaps even with a lot of money and resources. It is more dangerous because it is so powerful.

AI can be used in both good and evil. Let us use the new tools for the good of mankind. Above all, this includes enlightenment and transparency. Let us understand how existing regulations and laws can best be applied. Let us treat AI techniques in school lessons, educate the interested public. The more people know about AI, the better we can control it. New regulations only help to a limited extent.

By the way, the closer we get to the technological singularity in which AI systems become uncontrollable, the more we should conduct this debate on regulation. Perhaps then I will also change my mind.

1984, Alfred Nobel, And Artificial Intelligence

I had written that I want to make a positive contribution to this world and that I would take care of the spread of machine learning and artificial intelligence. Does that even fit together? What is the positive contribution of artificial intelligence today?

Amazon Alexa, Google Home, Apple Siri, self-propelled cars, intelligent websites are the harbingers of a wave of artificial intelligence that is just beginning. These harbingers are often criticized in the media. Is the permissive handling of our data justifiable? Without our data, without a lot of data, machine learning and artificial intelligence will not work. Are we prepared to pay this price?

If you want to deal with artificial intelligence, then I strongly recommend that you read George Orwell’s novel “1984”. Orwell wrote it in 1947/1948, it was published in 1949, and it is more topical than ever. “Big Brother is watching you!” In this novel we accompany Winston, who works at the Ministry of Truth and paraphrases newspaper articles, so that predicting the government is true in the archives, even if they do not arrive. And Winston, like all citizens, is watched day and night on telescreens in their homes and at work. The telescreens are reminiscent of devices like Amazon Alexa, which also listen in – and soon also observe, like the telescreens in the novel “1984”. What seemed like crazy science fiction for many years is now everyday life. Read the novel and get your own picture of the world we could create with artificial intelligence.

New technologies are neutral at first. Like a kitchen knife. It can be used for cooking, or it can hurt people. Good and evil are both possible. Alfred Nobel invented dynamite to do something good. He wanted to simplify mining. He did that, too. But then dynamite was also used to wage war. That made Nobel so difficult that he used his fortune to make the Nobel Prizes possible.

Every new technology is in this dichotomy. Since the Stone Age. The spear provided food when an animal was killed with it. And it was also used for fights between humans. The same sides exist with the steam engine, electricity, engines, airplanes, nuclear fission, whatever.

Machine learning and artificial intelligence are also able to make people’s lives much easier, autonomous vehicles will make roads safer and we will no longer waste time in traffic jams, but use them. The downside could be comprehensive monitoring and transparency of my data. Who knows.

I look forward to working with new technologies. I believe that in the end it will make the world a better place. “Technology as a force for good,”

said Pat Gelsinger, VMware CEO. That’s true. But we also have to reflect and consider ethical aspects. The ethics of new technologies is a field of its own, an important field. We have to be careful not to neglect this despite all the progress we have made.

That’s a core topic for my work at VMware.

My Path To Artificial Intelligence And Machine Learning

One of my personal goals is to make a positive contribution to this world. For me, this includes voluntary work, supporting family and friends, but also my job as an employee at VMware.

People spend money on things that improve their lives. Companies offer things. They can be products or services.

In this context, I look at artificial intelligence and machine learning (AI/ML). Can products and services be improved by AI/ML? Can AI/ML make a positive contribution to people’s lives? For many this is an open question. There are advocates and skeptics. I assume that the answer is “yes”.

Then the question follows: How can I help with my work at VMware?

VMware is a company that offers state-of-the-art IT infrastructures. IT infrastructures are the backbone of companies. This is where companies manage their products, customer data, processes and communications. And more and more companies are using IT infrastructures to improve their products or even offer completely new products and services. And these products or services should ideally improve people’s lives.

If a company wants to use AI/ML, it must first consider what it will be used for. What is the reason for using AI/ML? Then follows the consideration of which technology, which algorithms or which combination of technologies and algorithms are best suited for this requirement.

Then follows the determination of the AI/ML infrastructure. And at this point VMware can provide a very good foundation. Here I can make my positive contribution to the world by helping to connect the dots. The potential of VMware, the people, the companies, the technology, the ethics, and other aspects that are important when using AI/ML. It’s a complex field with many pendencies. AI/ML will conquer more and more areas. VMware’s contribution to the world’s IT infrastructure can be significant. I don’t know exactly what it will look like. It’s a journey with an uncertain outcome. But it is an exciting journey.

I will document my journey with the following contributions. I will raise questions, and maybe answer some questions. It is my personal view of the world of new technologies.