An interesting conference last evening in Geneva on applied AI, which confirmed my perception that we are no longer in the times of spectacular news, like winning different games against human players, but that we are talking now of challenges of practical implementations. A couple of insights:
Insight #1) The conference was organized by HEPIA, which stands for Engineering College in Geneva. (Note: you should not mix this with Geneva collège, which is the same like lycée in France or high school in the US). There was an impressive lineup of professors and researchers from this school talking about various aspects and challenges of implementing AI. So, an important insight is that you no longer need PhD to work on AI solutions. It is equally possible to do an AI bachelor’s or master’s degree from a university of applied sciences like HEPIA. This fact will have a huge impact on reducing costs and go-to-market time for most of the practical applications of AI.
This (and not not this) leads to my insight #2) Scientific work needs to become more agile and crossdisciplinary. There was a nice presentation of Prof. Nabil Abdennadher of their energy project, with the conclusion that the progress of AI will depend on its social acceptance. Which implies that AI research needs to be closely interconnected with social science. And in many cases not only social science, but many other domains. Another element comes to awareness with a recent study at Stanford : Why ideas get harder to find? which points out that the productivity of scientific work falls in all domains of research. In case of semiconductors the decline is 6.8% per year. Well, all this is a big problem and scientists have to find a solution for it. One solution track is : do agile research and get inspired with ideas of one of the leading thinkers in this field Alain Aspuru Guzik who works on highly automated, self-driving labs.
In the end, the opening speech by Pierre Modet was a bit disappointing. Geneva is falling behind other major cities in the adoption of AI, and this situation calls for more attention and support from the political circles, but I didn’t hear anything concrete in this sense. We are the global center of private wealth, commodity trading, trade finance, major international organizations and global governance, and there is no organized way of gathering use cases from these sectors and channeling them to be implemented with AI. And all this despite the fact that Switzerland is quickly becoming a startup nation. So, insight #3)Development of an AI ecosystem requires conscious and comprehensive efforts. We need the breakthrough innovations, but we also need the agile infrastructure and programs to convert these innovations into business and social value. And falling behind in this domain means loss of job opportunities for the young generations in the years to come and negative impact on economic growth.
Last month I had opportunity to drive some of the latest Mercedes models of different types through a rent-a-car agency. And they are very good to drive. But comparing with Tesla, not at the same level of experience.
Here is the list of my reasons to purchase one such full-electric car.
First, it is obvious for me that we all will soon have to give up from driving fuel-based vehicles and go with more sustainable and less-polluting cars. So, for me it was just a question which model to choose.
Second, Tesla is indisputably the best electric car. Comparing different options, we quickly see that German manufacturers wasted opportunity to lead in the development of this new generation of vehicles. There are new models , but only being marketed from this year. And I don’t like to buy the first generation of a product. Every new product brings some design or manufacturing problems and issues, which can be resolved only with the next design cycle. Tesla Model S was first produced in 2012, so the company had enough time to improve the original design and identify all major risks. And I appreciate very much this maturity. Then, there are numerous Chinese manufacturers and according to some recent market reports, China produces and buys more electric vehicles than the rest of the world combined. So, they have the experience and the technology. But here in Europe we don’t have opportunity to see and test cars like Geely Geometry A, BYD Denza or NEVS EV 9-3.
Third also, as I’m of Serbian origin, like the greatest scientist of all times Nikola Tesla, it has to be that I love this brand. And it is very nice to see that the name of this scientist who seems to be the only person who really understood our earth and its energy is being reestablished, especially after the proofs of recent times that Einstein’s theory of relativity can’t really explain the laws of universe.
So, my first impressions of Tesla Model 3.
You feel powerful. You feel the elegance of the white color and white interior, glass roof, and minimalist design. You enjoy in the silence, but the same time ,you know that you have an unparalleled Ferrari-level power of acceleration. I must say that one should use it very carefully, especially in the city where other drivers or cyclists don’t expect that you can jump so quickly from 0 to 50km/h. My Long Range model has 350 horse powers, all-wheel drive (4X4) and accelerates from 0 to 100 in 4 seconds. The Performance model is even more powerful and faster than this. The car is very very silent. And when you remove your foot from the accelerator, it activates generative braking which slows down the car more than traditional fuel-based vehicles. Something to get used to, but I like it. And while doing this, you hear a little sound like of a spaceship from SF movies. Wonderful!
Entertainment is fantastic. The car has a data SIM card and access to a Spotify account so now I can listen and discover music that I love, and simply add to my playlists. You can enjoy all kind of music (jazz, children, Bollywood etc). I’m so grateful that Tesla liberates me from stupid FM/DAB radio stations. But I can actually listen not only local, but also many world and foreign radio stations through the streaming service. The sound quality is so good that you actually prefer sitting there than in your living room. And the 38 cm touchscreen is large and very responsive to cover all dashboard information, driving visualizations, car commands, music and navigation map.
Apart from the touchscreen for the commands and dashboard, there are no other displays or buttons. Of course you have a lever to change from P to Drive or Reverse, and one for the turn signals and lights, and two buttons on thee steering wheel, and that’s all. The rest is available on the screen, the position of which is rather convenient. Where my hand was earlier to manually change the gears, I need to extend it only about 10 cm to reach the commands on the screen.
Fully Self-Driving (FSD) Capability is an option, which costs 6’300 CHF. I have just an autopilot, but FSD is easily upgradeable on the existing hardware. There are more and more self-driving features, and for more information on how this is implemented, check this video You will learn how sophisticated the chips, the neural networks and the datasets are so to enable this AI-powered self-driving car.
I plan to charge the battery mostly at home. Normal power socket can provide up charging of to 15 km/h, but the three-phase power enables up to 60 km/h. For my short distance rides, normal wall socket will be enough. The long range model with fully charged battery can go up to 500 km, so I think for my weekend trips or even holidays all this should be enough. And for longer trips there is network of superchargers to charge the battery in 25 minutes. The battery loses every day 1% of the energy, so that’s something one should also be aware of.
About the available space inside the car, it is interesting that under the front hood there is no engine. Instead, there is space for luggage. Rear section is spacious enough. In addition I also have installed a tow hitch for trailers of the weight up to 910 kg. Maybe not enough for the heavy and long campers for 4-6 persons, but it will do for the smaller ones.
Why Model 3 and not S or X ? Well, Model 3 is the most recent and the most technologically advanced. It is also smaller (4.7m long) and as I don’t have so much space in front of my house, it was a no-brainer I would choose this model. Dual motor for long range, with all options costs around 70’000 CHF. It is a considerable expense, but still affordable. And once you drive it, you can’t go back to the fuel cars. I think this was the idea : make a car not only sustainable but also enjoyable and available to the middle class.
With Tesla Model 3, and as of 2019, the history changes. Ordinary people like me can drive excellent and powerful electric cars, and the car manufacturing industry will see a lot of disruption in the next 5-10 years. I hope we’ll see more electric cars starting from this year, and not only cars, but also motorbikes – I think it is really pity that people here in Europe still drive petrol scooters and not the electric ones.
Does it mean that everyone should buy Tesla ? No, of course not, as there are many other brands (German, Chinese, etc). But think of the following fact: a normal fuel car produces around 200g of CO2 per km. For 10’000 km that people easily drive per year, it is 2 tons of CO2. And this is per car or per individual. As of July 2019, the measurements show that the atmosphere contains now 415 ppm of CO2, comparing to 390 ppm in 2010. And IPCC report states that 430 is the limit, above which we must not go if we want to be sure not to have a climate-related disaster, caused by more than 1.5 degree of the temperature increase.
I was reading this report from IBM Institute for Business value titled ‘Women, Leadership and the Priority Paradox‘ and it just came to my mind : how obvious and commonsensical this actually is.
Everybody should agree that intelligence is equally distributed among two genders. At least all of us could see that in the primary and secondary schools, where boys and girls are both able to understand math, learn languages, science and arts. Still, later in life, in the western and most of modern societies, men have more power and money and women less of it. Maybe you would say : “Men earn money to bring home and women decide how to spend it”. But who decides within these companies and governments how to spend money ? Following the same logic, then shouldn’t it be the women ?
I think not. I think that the power of decision needs to be balanced. And here I would come back to my argument that the intelligences are equally distributed among two genders. If this is so, we need to find a way how to use these intelligences in a balanced manner on the smaller, household level and the broader – corporate and national levels. In other words, the current situation where we have mostly men deciding on corporate and national strategies, and implementing these strategies is wrong and unsustainable.
In old times of early human history when humans lived as hunters and gatherers, genders were equal. This has been proven by the way how men and women were burried – no difference in the positions, sizes or the decorative objects that were found with their bodies. In the agricultural societies, at least in the early stage, women seem to be privileged. The evidence is female figurines of goddesses. Look at the art objects of ancient Vinca.
But later, all this changed. Gods became male, and with that, man became a kind of god. In the later history and through middle ages, kings and rulers were men, solders were men, priests were men, scholars were men, artists were men. What did they achieve? Definitely a lot. But also single-mindedness, which leads to conflicts and lack of tolerance and understanding. Conflicts that plague even the modern society. Shortsightedness and lack of broader perspective. It is like using only one eye. Even though you have two.
Many things changed in the meanwhile, but there is still a problem. Go now to any corporate or government office, and you will see that majority of managers are men. Think a moment about that.
Sustainability means long-term survival of the humans. If we want to achieve it, we need to change and broaden the current perspective. We need to open both eyes. We no longer live in the world where physical strength is so necessary to produce value. Machines and technology do all the heavy lifting. We control and manage them. There is no reason why most of the professional roles can’t be performed by women. And managerial too.
The benefit is very obvious when you look at Scandinavian countries, which managed to develop this balance. They are the cleanest, most modern, and most innovative countries with highest quality of life. Why ? Because they deploy the human intelligence where it fits the best. And both parents take responsibility of the childcare.
How to change the current situation in most of other countries ?
I’m thinking here of few practical steps.
First, governments should adopt legislation so that publicly listed companies must have gender balanced senior leadership and middle-level management. Private companies especially the smaller ones are simply private. But public companies have responsibility not only to the shareholders, but also to the society. So, they need to behave in a more responsible way.
Next, within companies, women need to get the chance to gain experience. They need exposure and this has to be formalized in the corporate policies. Similar like we all have to give up our combustion cars and either ride bicycles or get electric ones, we all have to accept that women should go higher in the corporate hierarchies. Even if the electric cars don’t have such a radius like the combustion ones, we have to do it, for the sake of our kids and our future. And even if these women are not as skilled as some senior male executives, they need to get this opportunity. If they don’t learn quickly enough to produce results, find other candidates. Collectively and in not so long-term , this will pay off.
On a personal and family level, I don’t have any particular advice as I’m confident that people who live together will help each other. It is beautiful to take care about children for both men and women. And it is also equally boring to clean the house to any of us.
I think it is the time for the situation to change. To get to the next level of our human development. To be more sustainable, and get in balance. To get in balance with the nature and among ourselves.
Disclosure: I have two daughters. And I want to see this world become a better place for them so they can be happy and successful in life. I can see in them so much strength, creativity and intelligence, and I’m teaching them that the sky is the limit, and not some glass ceilings on the way to it.
I was invited recently to take part in Blockchain Finance conference in Singapore as a panelist in a panel discussion about the use of blockchain for trade finance applications. I find this subject very interesting for many reasons that I will try to elaborate in this post. First I will just explain the main purpose and challenges of trade finance, and then what the benefit of blockchain in this particular field can be.
I wish to express the gratitude to the Blockchainer, Shanghai-based blockchain startup incubator with presence in various countries, for their invitation to this great event. The conference attracted 300 participants, and it was pleasure to learn the state of the blockchain finance sector in South-East Asia and China. There were many talks about blockchain payments, blockchain insurance, distributed funding, custodian services, cryptoexchanges, innovations in investing in blockchain projects, regulatory situation etc Beta Feng was great organizer of this event. Now, back to my topic.
What is trade finance ?
Trade finance means simply financing trading activities. Lets say a manufacturer or agriculture producer is making, or has already made a product and would like to sell it. Lets say that there is also a buyer that would like to buy, or maybe import to its country this product. The seller can invest money itself in the whole process, sign the sales contract with the buyer, load the goods on a truck or a ship, and wait for the buyer to confirm the acceptance and pay for the goods upon reception. The buyer can also decide to buy, sign the sales contract, wait for the goods to arrive, and pay the products from its own funds on at the time of reception of the goods. But often, they would decide to approach financial intermediaries (banks) to cover with a loan for this period while the goods are being produced, shipped and verified. This loan is one of the instruments of trade financing. The benefit of this loan is to provide the liquidity to the seller to buy the machines so that she can focus on making the product, improving the manufacturing, to perform the R&D etc. Using this financial instrument, the seller doesn’t need to wait for the goods to be shipped and received by the buyer. On another hand, the buyer doesn’t need to pay immediately for the goods. He can use the money for some more time (30 – 90 days) for the marketing promotion, to cover for logistics and warehousing costs, retail distribution. Trade finance instruments hence have a very important role in providing the bridge between the moment when the seller starts producing the goods and the moment when these goods are being paid for by final consumers. These instruments represent also a guarantee of payment, as seller and buyer very often don’t have history of mutual interactions to be able to trust each other So, the trade finance instruments allow all trade participants to focus on those activities they can perform the best: seller on production, buyer on distribution. In this way during the last 20 years, trade finance has stimulated, accelerated and expanded the globalization process to such extend that we can say that without these financial instruments there would have been no globalization today.
What are these financial instruments? Some of them are simple, like open account trading, advance payment, documentary credits and similar. Some are complex and include several participants and multiple control points. But in all of them, the intermediaries who provide this bridge funding will carefully manage risks. There are many kinds of risks, like manufacturing, product, transport, payment, and currency risks Most often, the financing providers will require the backing of the loan with the collateral, which in a simplest case is the goods itself being sold.
So, this is (very simplified) theory behind trade finance. It applies to commodity trading, or to trading of finished goods (machinery, consumer products, etc).
It is important to mention that global trade has always been very much influenced by technological progress and several innovations have particularly contributed to accelerate trade. Money is one of them. Money was invented to facilitate the trade, as before money the trade was limited to exchanging the goods between two participants. Money enabled people to understand the value of their goods and easily negotiate the price. Next, the standard-sized shipping container was the second most important invention, because it became possible to densely pack different products into these containers, stack large number of them on the ships, unload them quickly onto the trucks, and transport the goods efficiently to the final destination. Invention of large ships and barges made it also possible to transport huge quantities of goods per sea, the cheapest way of freight transport.
So, trading companies have always looked into the opportunities offered by technology to optimize and accelerate the trade process, in both logistics and financing the operations. And since the computers were invented, the trading companies were working on using them to gain efficiency. The big dream has been to automate the interactions between participants, but as of now, the trade transactions among sellers, buyers are their financing providers are still manually entered into their internal accounting systems, in many cases based on paper documents sent by DHL from one party to another. Automation of their internal processes has been achieved to a high degree, but the processes involving multiple parties are still manual, taking long time and suffering from many inefficiencies. One of the reasons of the lack of overall automation is (again) the mistrust between participants. There were many initiatives to introduce various digital platforms, but they until now have not been successful.
Now we have Blockchain
Two blockchain features are the most important when considering trade finance. First is distributed ledger. This means that if, let’s say, we create a network of 100 participants, each of them will have a full copy of all transactions on their computer. In case of private blockchains, one participant will see only those transactions described in the network’s rulebook where they are authorized to have access to. In case of public blockchains, all transactions will actually be accessible (private information will be hidden behind public keys, of course). The second feature is immutability. Modifying the previously committed transactions in the blockchain is technically very difficult and in some cases would be possible only through a hard fork where all participants agree to redo all transactions following the one that they want to change. It is not the same like in a centralized database, where a hacker or someone authorized by the management can change all data if they decide to do so. In blockchains, all participants must agree to this change and make conscious efforts to implement it.
These two important blockchain features in fact provide an environment in which the participants feel more confident in the reliability of the information they exchange and transactions they perform.
Of course, blockchain is not an all-cure medicine. If the data which enters the system is fake, and other participants are not aware of its incorrectness, blockchain will keep its record. So, validation of the trade documents and trade-related information will remain an important task to be done “off-chain”.
We should add here also automation through smart contracts. These are programmable blockchain scripts, which can perform certain tasks if they are invoked by the participants, or which can trigger in case of a particular transaction. Example would be: bill of lading is accepted by the validation organization, which will automatically trigger the generation of the payment request. The level of automation is of course configurable. For repetitive and well-understood operations, it can be fully automated, but for those requiring care and due diligence, smart contracts could be organized in such a way to involve human control points. Currently in this new world of distributed blockchain applications, an application could consist of hundreds of smart contracts that are enchained in order to implement very complex scenarios. The beauty of these applications is that they can invoke internal procedures and applications of various participants, can lookup external data resources like latest currency rates, or can use machine learning algorithms to perform classifications or identify the best course of action. And the end result, a new transaction or a record is included in a new block and accepted by all nodes in the network. So, we can have all three : automation, decentralization and immutability.
So far so good. The question really is if all this is sufficient in order to persuade typical sellers, buyers, and financial intermediaries of the trade finance process to trust the technology and trust each other, and really go forward in integrating more seamlessly interactions among themselves, which until now has not been the case.
It is worth mentioning that there are other solutions based on blockchain, which can complement nicely with the new trade finance platforms and applications. For example, blockchain-based payment systems, which aim to reduce the cost and automate the cross-border payments among the banks or among the end consumers. IBM’s Blockchain World Wire is one example of it. Another example is Chinese startup everiToken. There is also a set of solutions for the global supply chain, where participants can follow the flow of the goods. IBM’s TradeLens is an example of this as well. Many of these auxiliary solutions can add value to the trading process.
Hope this explanation on how blockchain supports the trade finance is rather easy to understand. Now the questions I got during the blockchain trade finance panel discussion were quite interesting, so I will comment also on them.
One was : “Are the consortium-based platforms the right way to promote the use of blockchain for trade finance, and are these consortia going to become a new norm?”
Indeed, many trading companies have recently started forming networks based on blockchain platforms. These networks are organized as consortia. Examples are We.Trade, eTradeConnect, Voltron, Marco Polo, Komgo and many others. It is enough to google these names to find quite a lot of latest information about their approaches and participants. Most of them use one of three underlying blockchain technologies: IBM-supported Hyperledger Fabric, R3 Corda or JP Morgan’s Quorum (which is based on Ethereum). I think this way of self-organizing, as of 2018 and 2019, is a very interesting way of experimenting with the technologies, learning how to benefit from them and how to automate the existing business rulebooks. Some companies are actually members of several networks, which allows them to compare the different concepts.
These networks will necessarily evolve, eventually more participants will come, more complex instruments will get implemented. I expect that the solutions will scale and the adoption will grow. As of April 2019, only a small number of transactions is being done through blockchain, and many of these are done in a parallel-run basis, alongside the traditional systems. I expect that it will take few years before the adoption grows to such a level that participants decide to move completely to the new blockchain-based solutions. In this sense, having three different platforms to play with is actually an asset. It also allows the technology providers to learn from each other and improve the existing solutions.
The second question was “what will be the role of the public chains in trade finance?” Very interesting question. Currently all trade finance is being done in B2B space, in particular among large companies. The problem is that smaller manufacturers and sellers from underdeveloped countries, even if they have capacity to produce and export interesting and good quality products, their funding applications are very often rejected by the banks who are not able to evaluate the risks, especially in the context of the still present consequences of the financial crisis from 2008. The current financing gap, as reported by WTO in their latest report, amounts to $1.5T. This is huge amount, which could in fact unlock the economic growth of many developing countries. The question for me is : can blockchain help to cover at least a part of this gap? I bet this is the role of public chains, with all financial instrumentation that they bring with them : tokens/ cryptocurrencies, cryptoexchanges, ICOs etc. Democratization of this process is a big promise, so I believe that there will be definitely a number of new solutions for this domain not covered by the existing intermediaries. I cannot comment on the details of the implementation of this kind of solutions, but there are many options and potential ideas.
At present, private and public blockchain platforms stimulate each other’s development much more than they compete, as they cover completely different market segments. As of now, trade finance participants (who are mostly large and regulated organizations) still view blockchain as an experimental platform which needs to prove itself, so they use it for a small proportion of their transactions. But as these participants gain experience and confidence, this will also stimulate the development of new solutions for smaller value loans, either by major platforms or by somebody completely new in this field. Optimistic scenario for me would be that in 3-5 years we see around 50% of the current world trade finance being performed on-chain with newly proposed B2B platforms, and that first solutions for the smaller value trade loans are being implemented on some kind of public chain.
The topic of blockchain for trade finance is very important, as trade finance is a pillar of the global world as we know it. Trade finance needs to be more efficient and more inclusive. As the shipping containers have indirectly enabled globalization, the world now needs a new set of solutions to further accelerate and expand the trading process.
There are many promising innovations : freight drones, hyperloop, unmanned ships, robotic container loading and unloading etc.But the most important innovations in trading and trade finance are expected to be in the domain of trust. And if blockchain can help achieve the needed trust among market participants from different parts of the world, in the situation of increasing complexities and difficult-to-understand risks, this will unlock the tremendous opportunities and a new phase of the human development. The impact is enormous. But it still needs to be seen if and how blockchain will actually make it possible to get there.
February 12 – 15 was the week for Think 2019, according to some analysis the most important tech event of the year. IBM assembled together all major individual exhibitions and this was the second edition of Think, now the main annual IBM conference. This time the event took place in San Francisco. I attended the event as one of the speakers on the subject of Watson Machine Learning and Watson Studio, and in my session I also had as guest Anyline, AI company from Austria.
The conference was successful and it was huge. Not only based on the number of participants (around 30’000), or sessions (more than 3’000), but also on the quality of discussions and topics. IBM events are not so much about announcements, but I will list and describe the most important ones in this post, along with highlights of the most recent and interesting technologies. Continue reading →
On January 31, I was instructor at the Quantum Master Class at IBM Studio in Zürich.
I coded in Qiskit (IBM’s Python SDK for quantum information science) and demonstrated how different Bell states can be implemented in jupyter notebook, how one can code the entanglement of three qubits, how to develop of the quantum circuit for bit addition, and the logic behind Deutsch’s algorithm with its implementation in Qiskit. Continue reading →
On November 9 this year, I gave a speech at the first international AI conference in Serbia. The venue took place in the beautiful city of Novi Sad, in participation of around 300 AI developers and enthusiasts, data scientists, businessmen and representatives of academia. A first thing that you realize when going to Eastern Europe for a tech conference, is a high level of technical interests and expertise of the local people. This was the case also in Serbia. When you talk to them, you realize they are serious developers. It is estimated that there are around 50’000 people in total working in different IT jobs in Serbia, and this region around Novi Sad has actually the highest concentration of skilled IT professionals. Continue reading →
It was fun to play with Watson Studio and create my own convolutional neural network. Watson Studio is complex but easy and graphically intuitive tool, and integrates well with Watson Machine Learning engine. I used standard MNIST data set (with images of hand-written digits), created three layers deep CNN, exported the algorithm in TensorFlow / Keras, trained the model both using GPUs on Watson and locally on my Jupyter Python Macbook environment and got nice results. To data scientists and AI developers : go and use it. And for all other enthusiasts, the best way to learn AI is to create it yourself 🙂 ! Continue reading →
Twenty years ago, IBM impressed the world with its specialized chess-playing system that won the match against the best chess player of all times, Garry Kasparov. Last year, Google repeated the same achievement with its Go-winning solutions AlphaGo, AlphaGo Zero and most recently AlphaZero. How did this evolution happen and what are the key success factors of these remarkable events in the history of AI ?
First, a little bit of history and background theory. The history of chess-playing AI solution goes back to 1957 when IBM research scientist Alex Bernstein wrote the first complete chess-playing program and ran it on IBM 704. But it was only 40 years later, in 1997 that computer-based solution reached such a level to surpass the best human performance. AI-based solutions for game-playing haven’t changed a lot during this period. The best practice was the use of opening books for the first dozen moves, and then use of Minimax algorithm for the search. Chess has up to 35 different moves, and can go to 50 moves deep, which makes it computationally impossible for any computer to identify all possibilities until the end of the game (35^50 is higher than the number of atoms in the universe). Due to the wide search, even nowadays this algorithm would go to the depth of maximum 10 levels within the allotted time for the move decision. Alpha-Beta pruning was developed to discard those subtrees that don’t contribute to the evaluation of the best moves for the Minimax algorithm. This helped to reach deeper levels (up to 18 levels from the current position). Once this depth is reached, an evaluation function must be defined to assess the strength of achieved positions and to grade the selected moves. Evaluation functions are very complex, and this is were humans typically excel. Garry Kasparov had for example extraordinary sophisticated evaluation capabilities, better than any other player. Continue reading →