Select Page
University of Warsaw and deepsense.ai join forces

University of Warsaw and deepsense.ai join forces

University of Warsaw researchers and deepsense.ai launch reinforcement learning project powered by Google’s TensorFlow Research Cloud

Researchers from the University of Warsaw, Google AI and deepsense.ai take on a new reinforcement learning challenge on Cloud TPU hardware accelerators. The goal of the experiment is to end-to-end train an artificial intelligence to play video games fully inside a computation graph.

A team from the University of Warsaw made up of Piotr Miłoś, Błażej Osiński and Henryk Michalewski, has started a collaboration on reinforcement learning research with Łukasz Kaiser from the Google Brain team and with researchers from deepsense.ai. This project is connected to a research program on RL that deepsense.ai started last year.

In the experiment, an artificial intelligence will be end-to-end trained to play video games fully inside a computation graph. Assuming that a game simulator would also be a part of the graph, this could make tasks such as training AI to play video games even faster than what deepsense.ai’s team achieved last year. The intent is to run the training process entirely on Cloud TPUs, which are new machine learning accelerators designed by Google. This will save time previously spent on communication between accelerators and a host computer.

The main experiments are being run on Cloud TPUs via the TensorFlow Research Cloud program and supported by Google Warsaw’s Antonio Gulli, Ignacy Kowalczyk, and Maciej Pytel, who are helping us to deploy our experiments on the Google Cloud Platform. TFRC provides ML researchers with access to second-generation Cloud TPUs, each of which provides 180 teraflops of machine learning acceleration.

Henryk Michalewski, a research team leader on the project, offered his appreciation. “Many thanks to Google for sharing early access to Cloud TPUs with us, as well as to Antonio Gulli and Ignacy Kowalczyk for providing the Google Cloud Platform power to deploy our experiments. With such a strong infrastructure, we’re perfectly equipped to tackle our ambitious goal and leverage the research on reinforcement learning efficiency we started last year.”

In 2017, Michalewski’s team from deepsense.ai published a paper describing a new method they had developed to train a robotic arm to grip a can of coke. Their work was recognized by the General Chairs of the Conference on Robot Learning (CoRL) as one of 11 noteworthy papers in the reinforcement learning and robotics category. The team presented the paper in November at Google’s headquarters in Mountain View. The method could be used, for example, to train humanoid robots to combine single steps into a walk or a run.

About deepsense.ai:
deepsense.ai supports organizations in unlocking their data potential at all stages of data science maturity. The company delivers AI-based solutions and helps clients grow data science skills in-house under the Training & Development Hub. It has also developed two products for data scientists: Neptune – the Machine Learning Lab, and Seahorse, a visual framework powered by Apache Spark.
deepsense.ai is founded and managed by CodiLime, an IT consulting and software engineering company providing professional services for global clients in three core areas: Networks, Security and Big Data & Data Science.

SpaceX Falcon Heavy Inaugural Flight

SpaceX Falcon Heavy Inaugural Flight

When Falcon Heavy lifts off in 2018, it will be the most powerful operational rocket in the world by a factor of two. With the ability to lift into orbit nearly 64 metric tons (141,000 lb) a mass greater than a 737 jetliner loaded with passengers, crew, luggage, and fuel Falcon Heavy can lift more than twice the payload of the next closest operational vehicle, the Delta IV Heavy, at one-third the cost. Falcon Heavy draws upon the proven heritage and reliability of Falcon 9.

Its first stage is composed of three Falcon 9 nine-engine cores whose 27 Merlin engines together generate more than 5 million pounds of thrust at liftoff, equal to approximately eighteen 747 aircraft. Only the Saturn V moon rocket, last flown in 1973, delivered more payload to orbit. Falcon Heavy was designed from the outset to carry humans into space and restores the possibility of flying missions with the crew to the Moon or Mars.

What is the ICO hard cap?

What is the ICO hard cap?

When looking to invest in cryptocurrencies, there are few initial factors to consider which could give you a good indication of whether or not a particular token has the potential to increase in value. One of these factors is the hard cap of the coin circulation. The hard cap is the total issuance of a coin that will be created. This amount is fixed and specified before an ICO is launched. The coin distribution can never exceed this amount.

Before looking at the importance of the hard cap, there are three distinct categories of coin distribution that’s often used in investing.

Circulating supply

This is the amount of coins that are actively being traded in the market. In some instances, projects could have all their tokens pre-mined and released all at once through the ICO. The circulating supply will, therefore, be the same as the maximum supply.

Other times, tokens have to be mined over time or coins are released on a schedule. In this case, the two supplies will differ. Either way, circulating supply tells you how many coins are doing the rounds at that very moment.

Total supply

This is the amount of coins that actually exist, including the ones that are not in circulation. Why would coins exist but not be in circulation? It could be for a number of reasons. A team might have mined coins but kept them back without putting it on the market. Or developers are required to hold their portion of the ICO for a set number of months which is to prevent coins being dumped as soon as it gets listed. The coins therefore exist but are not yet in circulation.

Maximum supply

This is also known as the hard cap and is the most important one of the three. It will tell you the maximum number of coins that will ever be created (Bitcoin’s hard cap is 21,000,000 tokens).  Not all cryptocurrencies have a hard cap. Ethereum, for example, has no maximum supply limit of ether.

Why is the hard cap important?

There are two main reasons to have an adequate hard cap.

The first has to do with scarcity. Think about diamonds for a second, we know the scarcer they become (i.e. there are only so many diamonds in existence) the more valuable they become. It’s the same with cryptocurrencies, simple supply and demand. If there’s a finite supply of a particular token, the value of the coin is likely to increase over time.

This, will in turn, protect the integrity and value of the underlying network. There is, however, a fine balance in getting this number right. Too low a hard cap and you won’t be able to raise enough funds to develop and grow the network.

On the other hand, if the network is flooded with tokens that don’t have a purpose because the hard cap is too high, the value of the coin will become diluted, causing a drop in the integrity and value of the network.

The second reason is closely connected to the project roadmap. For every amount raised, the startup should make sure there is a clear and concise purpose for it. Basically, if we raise so much, we plan to do this, if we raise more than that, here is where the funds will be applied.

However, we’ve seen projects raising hundreds of millions of dollars with no set objective of what to do with all that money. There have been instances of startups setting a funding target of $20 million but then go on to raise over $200 million because they didn’t set a hard cap. There’s no predetermined plan for the excess $180 million. If funds are not put to work effectively, especially within young startups, investors will never see their expected ROI.

 

That will be the beginning

That will be the beginning

“There will come a time when you believe everything is finished. That will be the beginning.” ~ Louis L’Amour

Iranian bitcoin-only e-commerce shop

Iranian bitcoin-only e-commerce shop

That year, an Iranian shoemaker with a bitcoin-only e-commerce shop called Persian Shoes reached out to the Bitcoin community. CEO Mor Roghani wrote on his website that he could not take other forms of payment to sell his products worldwide. He explained, “the problem is we operate in Iran and most payment systems either are not willing to serve us at all or impose a huge risk on our business.” While bitcoin has been helping him circumvent his payments problem, sanctions placed on the country by the US still make it illegal for US residents to purchase goods from the country.

Polish Tech Day 2017

Polish Tech Day 2017

 

We have a new partner!! Polish Tech Day 2017!
On June 16th, Google HQ in London is going to be the host of the next edition of Polish Tech Day. What will be the consequences of Brexit for the business relations between Poland and UK? What is the future of British financial sector and what can it learn from Polish pioneers of digital banking? How are the professional opportunities in CEE changing? You’ll learn the answers to these questions at Polish Tech Day in London, the largest event dedicated to Polish new technologies outside of Poland. Read more…

 

Micro-Meetup and Pitch-Off

Micro-Meetup and Pitch-Off

On July 4th, John Biggs will be holding a micro-meetup at Campus Warsaw. During the event, you will have 2 minutes to pitch in front of judges and then 2 minutes of questions. No slides, no props, needed.
The winner of the pitch-off will get a table at TechCrunch Disrupt San Francisco 2017 & the 2nd place winner will get two tickets to #DisruptSF.
The event is FREE but places are limited. make sure to RSVP here: https://micromeetup-pitchoff-wawa.eventbrite.com
You can sign up to pitch here (https://goo.gl/forms/eMdXZN4rRtoYS9KJ3) and we will only respond to the selected folks.

Freakonomics-A Rogue Economist Explores the Hidden Side

Freakonomics-A Rogue Economist Explores the Hidden Side

Started to read “Freakonomics”, totally recommend it (Buy at Amazon)

  • Which is more dangerous, a gun or a swimming pool?
  • What do schoolteachers and sumo wrestlers have in common?
  • How much do parents really matter?

These may not sound like typical questions for an economist to ask. But Steven D. Levitt is not a typical economist. He studies the riddles of everyday life—from cheating and crime to parenting and sports—and reaches conclusions that turn conventional wisdom on its head.

Freakonomics is a groundbreaking collaboration between Levitt and Stephen J. Dubner, an award-winning author and journalist. They set out to explore the inner workings of a crack gang, the truth about real estate agents, the secrets of the Ku Klux Klan, and much more.

Through forceful storytelling and wry insight, they show that economics is, at root, the study of incentives—how people get what they want or need, especially when other people want or need the same thing.

Numbers

Numbers

If you want new ideas, read old books.
“Grown-ups love figures… When you tell them, you’ve made a new friend they never ask you any questions about essential matters. They never say to you “What does his voice sound like? What games does he love best? Does he collect butterflies? ” Instead, they demand “How old is he? How much does he weigh? How much money does his father make? ” Only from these figures do they think they have learned anything about him.” Antoine de Saint-Exupéry, The Little Prince
If you take a look at the evaluation process of most startups, the only thing you can see is figures. Maybe they have to listen to the little prince and start asking questions about essential matters. What those essentials are to startups is what I don’t know. I just know that we chose the easiest path and mostly, the easiest is not the best. Would be a good idea if someone more intelligible than me would dig deep into this and come up with some better analytics, something beside figures, something that “The Little Price” would like reading…