prg.ai newsletter #40

The fortieth milestone issue of the prg․ai newsletter is packed with news and intriguing facts not only from the Prague AI scene. Keep reading so you don't miss out on anything!

🏡 What’s new at prg.ai

We educate public administration in AI

We have successfully trained over a hundred attendees at the Ministry of Regional Development. Our Lukáš Kačena presented the most important things from the field of AI, including practical exercises and examples illustrating its current applications in the public sector. Our longstanding collaboration with public administration continues to evolve, with each training session tailored to suit specific needs. This time, we partnered with EDIH CTU to further enhance the learning experience.

Lukáš addressed the significance of AI for Prague during the Developing the City for Industry and Innovation conference, which was held under the auspices of the Prague City Hall. The event primarily centered around the establishment of the Prague Science and Technology Park (VTP). Lukáš underscored the VTP’s role as a hub for education, a meeting ground, and potentially as a provider of more accessible regulatory sandboxes for testing purposes. You can watch the recording of the entire conference for more insights.

Introducing the Ten dos and dont’s

For this year’s Safer Internet Day, we contributed to the creation of the ten dos and dont’s for the use of artificial intelligence. The rise in identity theft and abuse, deepfakes and vice crime increasingly show the need to know the rules and risks of using AI. We have worked with the Ministry of Interior, CZ.NIC and Safer Internet Centre of the Czech Republic on a set of basic rules on how to use AI, which we introduced at an expert seminar on Safer Internet Day.

AI in healthcare

How is AI currently assisting in medicine, and what advancements can we anticipate in the future? From interpreting X-rays and monitoring heart rates with bracelets to expediting administrative processes and detecting early signs of Alzheimer’s disease, our Lukáš discussed these topics on Sama doma of Czech TV.

prg․ai Minor registration is opened!

The interdisciplinary university program, designed for all students passionate about AI, is now accepting applications for the next academic year. This program offers a unique collaboration of top teachers from four faculties, providing the best courses in artificial intelligence under one umbrella. Don’t miss out on the chance to connect with key players in the Czech AI and IT scene and become an integral part of our community. Register by March 21!

February Industry Talk on catching fraudsters with AI

Regular Industry Talks are an integral part of our inter-university programme prg․ai Minor. The most recent one was held by Resistant AI, who focus on cybersecurity in finance. Ladislav Maleček, Samuel Mešša and Tereza Štefková presented to the students how they leverage machine learning to detect fraudsters, as well as the latest trends and challenges in combating money laundering.  We cherish the connection between students and practical application. Thanks to Industry Talks, students have the opportunity to engage with local companies and experts in the field of AI.

New members on board!

We welcome ententee and Profinit among our members! ententee is a development company specializing in the creation of intelligent applications across various industries, utilizing AI techniques like expert systems and machine learning. Their solutions are deployed in hospitals, trains, engineering firms, and online newspapers, addressing tasks ranging from patient tracking to content generation. Profinit offers consulting services in data science, machine learning, and AI. They collaborate closely with leading academic institutions, deliver lectures, and support educational initiatives such as Vilekula. Additionally, they actively contribute to community building efforts and provide public datasets for educational purposes, thereby playing a crucial role in popularizing ML and AI.

AI and language in Wired

Our collaboration with Wired has generated quite a number of commentaries. The next imaginary notch is from Katerina Lesch, who has focused on forensic linguistics. In the commentary, you can learn, for example, how to distinguish AI-generated text from human-written text, or what are the main linguistic characteristics observed in authors in forensic investigations. Can we actually distinguish whether a text is produced by a human or a machine?

Where can you meet us?

The incorporation of modern technologies into public offices or the integration of AI in shaping public administration will be the key points of discussion at the AI in Public Administration conference, taking place on April 4th and 5th in Hluboká nad Vltavou. Experts from the whole country from various fields will explore strategies for elevating authorities to the next level. Among them will be our very own Lukáš. Don’t miss this opportunity to engage in discussions with representatives from local governments, the public sector, universities, and companies involved in pioneering new technologies for public administration.

As a community partner, we invite you to the 9th edition of Machine Learning Prague 2024 taking place from 22 to 24 April. Join us to explore topics such as overcoming reasoning limitations of large language models, utilizing machine learning for exoplanet discovery, and constructing recommender systems using LLMs, among other practical talks. Moreover, don’t forget to use the code “prgai” to get a 15 % discount on your ticket purchase. If you’re still hesitant to come, tune in to Jiri Materna on the Data Talk podcast (Spotify/Apple), where he delves into discussions about the Machine Learning Prague conference, alongside other engaging topics.

This year we are partnering with WebExpo 2024 for developers, designers, product managers and anyone with an interest in the web. May will be here sooner than you think and tickets are disappearing. It would be a shame not to use our discount code “PRAGUEAI” with which you will get a 20 % discount on your ticket purchase.

👏 Local AI scene news that makes us proud

Misuse of AI in elections

Michal Pěchouček from AIC FEE CTU spoke in Studio ČT24 [14:45] about the influence of AI on the upcoming elections not only in the USA. Among other things, he discussed disinformation, deepfakes and content moderation on social media.

Realistic videos generated by AI

Sora’s new AI model generates realistic videos that are almost indistinguishable from normal ones. Viliam Lisý from AIC FEE CTU commented on the innovation for Studio 6, while Jiří Materna, founder of MLC and David Černý from AV ČR commented for Události.

Politically impartial AI?

How politically impartial is AI? And in what ways can EU regulation be a threat? Tomáš Mikolov was a guest of Čestmír Strakatý, where he reflects on his work at Google, Microsoft and Facebook, as well as his return to Czech science.

Easier planning of surgeries

A new algorithm developed at FEE CTU will simplify the planning of surgeries. According to ČRo, it will take into account the length of the surgery, the capabilities of surgeons and threats, and will find the best option even in the case of an acute surgery.

IT evenings

Want to be prepared for what’s ahead in AI in 2024? FIT CTU hosted a lecture as part of the Informatics Evenings series with Pavel Kordík and you can now watch it on record.

Students as astronauts

A new virtual reality app has been developed to simulate spaceflight, life in orbit and the risks to astronauts, aiming to promote space topics in educational institutions. Scientists from the FEE CTU in cooperation with Deloitte and under the auspices of Planetum – Observatory and Planetarium Prague participated in the development.

Innovation in industrial production and logistics

The ROBOPROX project, with a budget of CZK 467.9 million, was selected under OPJAK. It centers around the utilization of algorithms in production planning, enhancing energy efficiency, and deploying autonomous robots and drones. The project engages researchers and roboticists from FEE CTU, FIT CTU, and CIIRC CTU, with CIIRC CTU serving as the project coordinator.

Will a humanoid robot hang laundry for us?

Solving a Rubik’s cube or playing chess might be a complicated task for many humans, but it’s trivial for a humanoid robot. However, it struggles with everyday tasks such as climbing stairs, hanging laundry, or washing dishes. Tune in to Tomáš Svoboda from FEL CTU as he discusses the challenges of developing humanoid robots.

Will we change public transport in Texas?

Citya is collaborating with the Faculty of Transportation at CTU to implement an operational mass transit project into the American transportation system. The plan involves utilizing minibuses and a smart algorithm to create virtual stops near passengers’ current locations.

No more fake discounts!

Apify is helping the European Commission in combating fake discounts. By utilizing its Price Reduction Tool, Apify checks prices on over 700 e-shops, downloading price data daily to identify whether a particular e-shop is cheating customers with fake discounts.

A successful year for RecombeeLab 

Last year was a very productive year for RecombeeLab, the joint research laboratory of Recombee and FIT CTU. They published a number of groundbreaking research papers in the area of recommendation systems in leading journals such as IEEE Transactions on Neural Networks and Learning Systems and Expert Systems with Applications, or at the prestigious AAAI conference on Artificial Intelligence. Check out a summary of their best work.

The pitfalls of AI-generated phone calls

The US is combatting fraudulent phone calls wherein AI impersonates celebrities. Josef Vopička from MAMA AI provided insights on this issue for Studio ČT24 [15:25]. He discusses the mechanics behind AI-generated phone calls, methods to identify such calls, and anticipates what lies ahead.

Document authentication and cross-border payments

Resistant AI is providing their Document Forensics solution to Kooperativa. This technology will be used to verify documents submitted as part of Kooperativa’s claims process. The same solution is also used by Verto to verify documents required for cross-border transactions.

How to make data visualizations?

Creating quality visualizations requires careful manipulation of data, selection of appropriate tools, and often a long process of fine-tuning various parameters. Datamole introduces the creation of effective data visualizations as an integral part of Exploratory Data Analysis (EDA).

How to present data to a wider audience

An article from Datamole delves into the challenges associated with presenting data to a non-expert audience and offers practical tips on overcoming them. Leveraging real-world experience and examples of unsuccessful data visualizations, it sheds light on common mistakes and pitfalls encountered when presenting data.

No more paperwork with the new AI

Rossum has introduced a new AI model, Aurora, that efficiently processes invoices and contracts. It’s faster and more reliable than previous systems, plus it helps with political data analysis.

The MAMA AI Academy

There are still places available in the regular The MAMA AI Academy on 19 and 20 March. Don’t miss this ideal opportunity to learn about AI in a hands-on way. Focused on Business AI and document processing.

Caelestinus Digital Health Incubator 2024

The third batch of the Cealestinus Incubator, which focuses on solving interoperability challenges in digital health, has kicked off. This year, 22 startups focused on innovation in healthcare and health technology are participating.

Qminers reinforcements

Qminers is strengthening its ranks. Among the incoming experts is the new CTO Filip Matzner from the AI app startup Iterait.

AI helps therapists

The platform of the Czech-American startup Upheal offers assistance with writing session notes, aiming to support mental health professionals who often struggle with time constraints, overwork, and burnout. Upheal provides solutions to alleviate therapists’ administrative burdens and conserve their time and energy.

🎭 Events we’re excited about

8 Mar – Become an AI expert for a day – Have you ever controlled a robodog? High school students have the opportunity to experience working in the robotics lab at FEE CTU.

9 Mar#AimtechHackathon TechTalks – 3D manufacturing, AI in humanitarian aid, organ exchange and healthcare or speech synthesis.

11  Mar AI Safety seminars – Another round of AI Safety seminars for all AI enthusiasts. You can choose from a seminar focused on AI governance or a technical seminar on AI safety. Prague/Brno/online

11 Mar Creative AI Meet-Up #5 – Next meetup on the topic of video generated by AI. A hands-on event full of experiments, so don’t forget your laptop at home!

11 – 17 MarBrain Week – A festival of the latest discoveries and trends in brain research and neuroscience, part of Brain Awareness Week (BAW).

13 MarCharles Translator – Martin Popel’s seminar will delve into the mechanics of automatic translators, showcasing a translator developed at MFF UK as an illustrative example.

14 MarAI and creativity – “Will generative AI surpass us, and does it even make sense to be creative?” – This thought-provoking question will be explored in the next meetup of AI Horizonty, a series of debates featuring experts from various fields.

19 MarBusiness Breakfast – Traditional KPMG breakfast, this time on Responsible AI: AI Act

20 Mar Digitising SME 2024: Using AI in Industrial Manufacturing – Case studies, experiences and lectures on AI in manufacturing. For all those interested in automation and digitalization.

21 Mar Global Goals Summit – Petr Kadera, Head of the Intelligent Systems Group at CIIRC CTU will be one of the speakers. He will focus not only on sustainability and the use of AI.

21 MarAI in Context: Aligning with Human Values – Another seminar on AI in the context of different disciplines. Presented by Jan Kulveit.

22 MarTechnology Olympiad 2024 – AI, electromobility and cybersecurity. High school students compete in the most current trends for a prize of CZK 30,000 and a ticket to FEE CTU.

26 MarCOFIT Career Fair – The most interesting IT companies in one place. The job fair is suitable for students of all years and study specializations.

27 MarArtificial Intelligence in Video Games – Seminar by Jiří Hanák focused on video game development.

2 AprTruesdays – regular meetups for tech & business – every first Tuesday of the month.

8 AprAI for Nonprofits – How AI can help nonprofit organizations, simple techniques for implementing AI, and examples from practice and hands-on exercises.

10 AprAI for document automation (and how to do it without ChatGPT) – Martin Holecek’s seminar

You can find all the events we find interesting in our Google Calendar.

🧩 Stuff that interests us

Do we trust AI’s recommendations about our health?

Researchers from Masaryk University studied the usage of AI among Czechs in relation to health. They discovered that almost a third of people have utilized AI in some capacity. While the use of AI varied, trust in its recommendations regarding health was found to be low.

Generative AI in the public sector

Are we facing a revolution in public services? Generative AI could increase productivity and reduce bureaucracy. The Alan Turing Institute has conducted a major study on the use of generative AI in the public sector. This showed, among other things, that only 22% of employees are actively using it.

Something from every AI industry in one place

With more than 900 sessions, over 20 technical workshops and a variety of networking sessions, you can look forward to it all at the 2024 GTC AI Conference. Join industry leaders, developers, researchers and strategists who are shaping the future of AI and accelerated computing. And you won’t be in California from March 17-21? No matter, most of the sessions are streamed online.

How to build a successful startup

Central Europe is becoming a global technology hub and many startups are being founded and based here. But how to build a successful startup here? And what about the burning question of whether to stay or relocate to another market? Find out what the pros and cons of relocation are and much more at NVIDIA’s Beyond Borders online seminar on March 19.

A need for regulation of Open Foundation Models?

GPT-4 and LLAMA 2 innovate the field of AI, raising questions about their transparency, workflows and the need for regulation. Researchers at the Stanford Institute for Human-centered Artificial Intelligence have addressed what such regulation should look like, emphasizing the need to respect the needs of developers of both open and closed models. The key should be a balance that minimises risk without inhibiting innovation.

📢 Community calls

Google Summer of Code 2024

The Stratosphere Lab at AIC FEE CTU has once again been accepted into the prestigious Google Summer of Code 2024 program, providing bright programming students and those interested in developing cybersecurity technologies with a unique opportunity to implement their own ideas under the guidance of experienced researchers. Registration will be open from 18 March to 2 April.

Summer School at FIT CTU – With Artificial Intelligence in the Online Space

FIT CTU and AI dětem have opened registration for the summer school – With Artificial Intelligence in the Online Space. The event takes place from 22 to 26 July and is intended for high school students who are interested in modern technologies and want to use artificial intelligence tools effectively. Interested participants do not need to have any initial IT knowledge.  Applications are due by 15 March.

💼 Open job positions

Artificial Intelligence Center FEE CTU

Apify

Citya

Datamole

DataSentics

DNAi

Gen

MAMA AI

Qminers

Recombee

Resistant AI

Rossum

EY

KPMG

Valeo

Technologická gramotnost