Peter Metzinger

"#AI in #healthcare - opportunities, challenges and solutions" was the title of the session that Health Hub Vienna, LISA - Life Science Austria, INiTS | Vienna's High-Tech Incubator, EIT Health Austria held. It was, rightly, super overbooked. Georg Langs, with his contextflow and his Medizinische Universität Wien hats on, inspired us all with the way AI is affecting and will affect healthcare, from predicting disease course, responses to treatment, optimizing patient management, enabling secondary but also primary prevention (I will come back to that point), as it is already happening, but also to develop new therapies by identifying new early endpoints, incorporating patient satisfaction and QALYs, discovering new therapeutic targets, identifying new phenotypes, among others. In the last few years, there has been a shift from AI to support complex cases to AI to support healthcare efficiencies (starting to be of interest for payers for this reason at least), and narrow or close the healthcare workforce gap. Andreas Strouhal represented those many doctors that want to do their best for their patients, want to be able to recall and connect the dots from what they have read or heard on papers and conferences, and make it available at the point of care, right there, for the patient. AI not only to make the system more efficient and be able to see more patients per minute, per hour, but to be able to connect truly with them, to look at them in the eye, to put a hand on their shoulder. AI to augment and not to replace doctors. Helene Prenner, from ELGA GmbH but also in her own name, started with one of these statements that mark pitches "Data has no boundaries" and talked us through the challenges Europe is going through to make the most out of the fragmented data and use it for good, for individual patients, but also for research and to advance healthcare and society. Endri Deliu, bringing the US perspective, with a hopeful message Europe, said loud and clear to move faster, to try bold solutions, think much bigger, embrace change and failure and collaborate. He said we need fuel: talent, which Europe and Austria has, and money, and not "old-money" -as he called banks and insurance companies-, new money that invests and bets with Venture arms to speed up things and build the muscle. Regulation, -of course on the table-, may bring inequalities in the adoption of AI, bringing giants ahead of SMEs as they can afford employing people to make sure they comply with every line. To the question of using AI to shift from a sick-focused system to a health system, Helena replied with a call to collaboration, accounting for social-determinants of health. Irene Fialka, marvelously moderated the session, gave me the opportunity to ask a final question: how can we shape the present of the workforce challenge and attract the right AI and EI equip talent to the healthcare workforce for a sustainable future. Anita Gondola and team rocked!! #ViennaUp

39

4 Kommentare

Scoop: Statistisches Bundesamt (Destatis) has suspended some of its most important indicators after botching a data update, leaving citizens and economists in the dark at a time when the country is trying to boost flagging growth. In a nation once famed for its punctuality and reliability, even its notoriously diligent beancounters have become part of a growing perception that “nothing works any more” as Germans moan about delayed trains, derelict roads and bridges, and widespread staff shortages. https://lnkd.in/eurDGHdy

24

1 Kommentar

Switzerland needs a Science Advice Mechanism. The pandemic has shown how important continuous cooperation between science and politics is to enable evidence-based decisions in crisis situations. To this end, I propose a structure that is clearly anchored in politics and in the academic world and also integrates civil society. More on this in my upcoming book. #science #politics #publicadministration #verwaltungen

40

4 Kommentare

Good news! The complains by EU stakeholders that the #deadline for providing feedback for the #GPAI #CodesofPractice is too short have been taken into account. The new deadline gives #companies, #academics and #civilsociety eight additional days to finalize their contributions. Doing public consultations over the summer break is still not great and even the extension will not help everyone to meet the deadline. However, one should consider that: ➡ The new deadline is roughly in line with the 2021 Better Regulation communication (Chapter 3, page 5). I have recommended an extension until 20 September in my LinkedIn post in July (https://lnkd.in/eEsVc-vs). ➡ The timeline for drafting the #CoP is incredible tight due to the deadlines set out by Art 56 and 113. A further extension is barely possible if we want to execute - at the same time - a multi-stakeholder approach that allows all interested parties to jump in at different stages of the drafting process. By the way: this means that you will have additional chances to contribute. ➡ The whole episode underlines that public scrutiny + constant feedback can make a difference. The European Commission has again adjusted its position after realizing that there is room for improvement. Let's focus on improving the process more and more! 🙌 🚀

98

3 Kommentare

The relevant commission of the Swiss National Council has today decided not to pursue a simple, fair, and effective emission pricing solution - based on two weird reasons. As per the relevant press release (currently only available in German and French, https://lnkd.in/dFWfkFtM): 1. A majority of voters would not support a 'systemic change in climate policy'. But: They have no clue if that is the case. It has not been tested. And given that some 70% of the population (incl. all the lower income groups) would experience a net financial gain, IMHO it is quite likely that a well run, well communicated campaign will indeed gain majority support. 2. A 'strict enforcement of the polluter pays principles would also lead to competitive disadvantages'. What? That a worry for Switzerland, a service-based economy, one of the richest in the world, blessed by carbon-free hydro power and a long standing global leadership in both GDP and innovation per capita? Isn't the point of pricing pollution to hurt the competitiveness of polluters? Yep, they are out there, so obviously not everyone will come out ahead?! The good news: Today's decision has been tight (12 votes opposing, 11 supporting, with 2 abstentions). The thing to do now is to look ahead and keep fighting. Starting point for that is December 12 in Geneva, when we will hold a roundtable event on how to get the right pricing policy implemented. It is held as part of the Building Bridges conference. Registration is free for conference attendees, but by invitation only. Those interested to attend may contact my colleague Juan, using the detail listed on the program.

25

14 Kommentare

Difficult for European companies in the scope of EUDR to navigate such unstable legislative context ! Let's hope for more clarity on the final rules & implementation before end 2024 Coraline Divoire

2

1 Kommentar

Did you know you can have #AI #ArtificialIntelligence extract information from a pdf and create a podcast that talks about the content so that 'the layman' can understand scientific papers? For free? You can try Google Illiuminate here: https://lnkd.in/gNAJXUQA "Illuminate" is very restricted right now (only pdf hosted by certain domains allowed). But "Notebook LLM" accepts pdf's, websites, YouTube videos, audio files, Google Docs, or Google Slides. NotebookLM will summarize the content. You can ask questions. And it can auto-generate a podcast. How cool is that? Try it here: https://notebooklm.google/ Of course, people already auto-created podcasts about the dangers of AI. And jail-broken Notebook LLM to have the two AI speakers use plenty of swearwords...

3

1 Kommentar

According to Science|Business, the #EuropeanCommission is considering funding for #digital #research, starting in 2025, and covers #AI, #quantum, #photonics, #chips, #networking and related fields. In all, the proposed calls would add up to €550 million, as part of a larger “cluster” of #industrial #technologies in #Horizon #Europe.  Some of the calls under consideration are big, some small. One would set up a previously promised “hub” to interconnect European AI developers, as part of the GenAI4EU initiative. That initiative, announced as part of a big AI policy push last year by Commission President Ursula von der Leyen, is supposed to develop novel uses of AI in 14 industrial sectors. The hub is intended to interconnect everybody involved in that, carry out a “comprehensive landscape analysis of generative AI initiatives,” and “support the uptake” of the resulting GenAI4EU solutions in other parts of Horizon Europe. It’s also supposed to “support the European Commission’s AI Office in its function to promote an innovative ecosystem of trustworthy AI.” The budget: €3 million, a relatively modest sum by tech industry standards. https://lnkd.in/dcFTEB-V

5

COLUMN: The new Swiss report on commodity speculation reads like an all-nighter effort by a first-year college student — or even something spat out by ChatGPT when asked to “write a report on commodity speculation in the dull manner of the Swiss Federal Council. Don’t conduct any original research; ignore any criticism of the industry; and above all, do not name names.” The result is more interesting for what it doesn’t say than for what it says. Absent are, for example, the very names of the biggest players in the industry — all with major operations in or around Geneva: Cargill Inc., Bunge Global SA, Archer-Daniels-Midland Co., Louis Dreyfus Co., and China’s Cofco International Ltd. Absent, too, is discussion of the concentration of trading firepower among them and any details about how much money they made during the most recent price spikes. If one is investigating the agricultural-commodity industry and doesn’t know how much money the biggest players made, well, then, they don’t know anything. Rather than conducting its own research, the Swiss study analyzes the conclusions of other reports. Trouble is, some of those reports were already meta-analysis — analyses of analyses. So, ultimately, the Swiss spent two years to assessing others’ research. 

160

10 Kommentare

"AI needs the non-quants too", says Politico. In a new policy brief from Data & Society Research Institute, Serena Oduro and Tamara Kneese, Ph.D. argue that "AI Governance Needs Sociotechnical Expertise". They make the case for expertise from the humanities and social sciences being front and center in governance design, deployment and evaluation -- particularly in the current "AI talent surge" happening across the federal government. And bonus! Check out our friends at the Center for Democracy & Technology: Miranda Bogen and Amy Winecoff just released the related, "Applying Sociotechnical Approaches to AI Governance in Practice". https://lnkd.in/em2NFBep

237

7 Kommentare

🛰 Keen to propose a novel #geospatial #foundation model - not for one, not for two - but for entire seven!!! - problem tasks? 😎 🎉 Looking for a short-termed coding snack in-between the major European Space Agency - ESA ESA Phi-Lab Austria challenges? 👉 Then join the #AIforEarthChallenge 2024! Clay is challenging the #AI4EO community: Dive into #landcover #classification, #aquacultural #detection, and more during this exciting challenge. Attention, only open until 09 September 2024 so be quick! 🗺 #Landcover classification [Spectral, Classification] 🌊 #Aquaculture detection [Spatial, Classification via Similarity Search]  🌳 #Carbon above ground stock [Spectral, Regression] 🌾 #Crop yield estimation [Temporal, Regression]  ☔ Disaster response #floods [Spectral, Change detection] 🔥 Disaster #fire severity [Temporal, Change detection] 🌫 #Cloud Gap Imputation [Temporal, Generative] #AIforEO #MMFM (cc Cristian Bodnar)

41

🤖 #Systemic #resilience meets #genAI 🙌 What happens when #assetleveldata is combined with the #physicalrisks and #naturebasedsolutions in the form of #embeddings? 🧐 📢 The final call to the advanced #AI community - #climate & #nature need us! If you are passionate about complex math and want to: 💡 Design your #geospatial #foundation decision-maker with the minimal hallucinations; 💡 Embark on the exciting journey of building high-performing #climate and #nature #resilience models; 💡 Undertand how to tailor your #environmental #copilot for public and private #finance and #governments - then this #AIforEarthChallenge is definitely for you 🤗 ⏱️ The countdown to submit the best (pioneering!) #foundation model for #resilience #analytics begins! Try to beat us in proposing the most comprehensive synergistic solution to tackle global #systemic issues, using 7-in-1 lightweight generator 🤓😬🙌 🌳 Imagine one model agile enough to predict #climate #hazards - and propose #naturebasedsolutions to tackle those, on the fly? We aren’t far off, so get on the Cloud and finish your Code 👩‍💻 Clay ESA Phi-Lab Austria #NbS #SystemicResilience #NatureFinance #ClimateFinance #FairZero Cambridge Judge Business School Green Finance Institute United Nations Office for Disaster Risk Reduction (UNDRR) Google Cloud

12

🤔 Yesterday I provided testimony to the German Bundestag's Public Hearing on "Innovative Data Policy: Potential and Challenges" 🔹 Main Points: 📊Data, when used and reused responsibly, holds immense potential for advancing the public good. 🌍 Data policy, till today, has mainly focused on preventing the misuse of data (for good reasons) without preventing missed uses of data ➡️ Key area for policy innovation involves efforts to accelerate access to high-quality, timely data for re-use by public sector and civil society. 💡Among the innovations in Data Policies I proposed: 👉 Embrace Digital Self-Determination: Empowering individuals beyond consent, with innovations like 'data assemblies' 🗣️ and 'social licenses' 📜. 👉 Strengthen the public demand for data by identifying 📊 (100) high-value questions for which data needs to be made accessible; 👉 Establish roles like 'chief data stewards' 📈 that can enable data access in a more systematic, sustainable and responsible manner. 👉 Support the creation of trusted data intermediaries and new 🤝 data-commons models. 👉 Democratize access to high-performance computing 💻 for public interest purposes. 👉 Launch a Data Innovation Fund 💰to support new data institutions and initiatives 👉 Mandate 📝'data impact assessments' in policy initiatives 👉 Develop decision accelerator labs 🧪 to translate data into decision intelligence 🙏 Many thanks to the chairperson Tabea Rößner for the opportunity to participate, and to the other experts for sharing their most interesting perspectives: Aline Blankertz, Christoph Lange-Bever, Elisa Lindinger, Max Schrems, Prof. Dr. Rolf Schwartmann, Rebekka Weiß, LL.M., @Ulrich Kelber and David Schoenwerth 💻 All testimonies and summaries at 🔗https://lnkd.in/e7tFbZPS #DataPolicy #PublicGood #Innovation #DataStewardship

78

6 Kommentare

Ambitious climate policies are not just necessary but also economically viable. #kudos Lukas Bretschger ETH Zürich! for 'New Approaches in Climate Policy: Economic Pathways to Sustainable Growth' Achieving our climate goals requires innovative and effective policies. A recent article by CEPR - Centre for Economic Policy Research presents "economic pathways" as a novel, theory-based approach to climate policy. Here are the key takeaways: 🔍 Economic Pathways: These are dynamic models that simulate different policy impacts on the economy, providing a clear visualization of the consequences of various climate actions. 🌿 Decarbonization: The research shows that deep decarbonization only modestly delays economic development. In fact, stringent climate policies can ultimately lead to higher medium and long-term consumption growth. 📈 Policy Scenarios: The study explores three scenarios - business-as-usual (BAU), standard, and super escalator models. The super escalator, with a progressively increasing carbon price, is the most effective, achieving full decarbonization by 2040 with minimal economic disruption. 👥 Population Impact: Interestingly, population growth does not significantly interfere with economic development or resource use within these pathways. 💡 Policy Implications: The findings underscore the importance of robust, progressively stringent climate policies. These not only mitigate emissions but also support sustainable economic growth. ✊ Let's push for policies that secure a sustainable and prosperous future! Stay tuned for the best pricing policy #climateincentive. 🔗 Read the full article here: https://lnkd.in/dK44cX5z #ClimatePolicy #ClimateAction #Sustainability #EconomicGrowth #Decarbonization #RenewableEnergy #FutureGrowth

6

It's Blockhain Time! See you next week at the annual Austrian Blockchain Conference (https://lnkd.in/diDVEF_a) or one of the other events of the Vienna Blockchain Week (https://lnkd.in/dF8Y6JJx) Thanks to all who made this happen; especially Walter Moesenbacher from ABC Research and Ed Prinz from DLT Austria #blockchain #event #awesome ABC Research

18

The recently created European AI Office (https://lnkd.in/e-SMsmZT) may have a tough time finding the necessary technically competent civil servants to implement and enforce the #AIAct. An analysis of LinkedIn data reveals that the European Commission employs 38,285 people. A breakdown of their proficiency in modern skill sets gives the following numbers: Data Science - 356, Machine Learning - 665, Technology - 1,596, Strategy - 6,040, Policy - 11,026, and Project Management - 11,547. Additionally, it appears that most EU personnel skilled in AI are engaged in projects from the Horizon Europe research programme or are part-time contributing experts rather than full-time functionaries. A company such as Goldman Sachs has approximately 3 times the proportion of data science professionals compared to the European Commission and about 2 times the proportion of machine learning professionals.

16

3 Kommentare

In a system designed for their benefit, #monopoly utilities overestimate demand, leading to overbuilding the grid, leading to over-collecting profits from captive ratepayers. @PattyforGaPSC captures all the flaws in one excellent analysis: https://vist.ly/3g7xx

6

1 Kommentar

There are plenty of touchpoints between #AI and ethics. Much of the public discussion centers on questions of existential risk, of algorithmic bias, surveillance, privacy etc. 💰💸🤑 What is hardly ever discussed is: Contrary to other ground-breaking and potentially harmful technologies often developed in universities, the most powerful AI applications stem primarily from private corporations driven by #profit. ➡ This means that questions of AI ethics must always be linked to #business ethics, and its core elements like corporate #responsibility, #accountability along the value chain and towards #stakeholders, and safe and responsible products. 📰 In a recent article for the magazine "THINK", published by The HEAD Foundation, I discuss how well AI firms are adapting to current day notions of business ethics (link in comments). 🤫 Spoiler: poorly! Leading AI companies' singular focus on money and power leaves little room to consider the rights of those who contribute to and are affected by AI advancements. Put differently, it is at odds with modern notions of corporate responsibility. 👻 the AI industry tends to overlook supply chain considerations and fails to acknowledge stakeholders: AI companies like to pretend, and users want to believe, that their products are mainly virtual and intangible and that the work behind them is either done by #algorithms or by highly paid human engineers. Yet, as evident by the term "Ghost Work" this is not true. In particular, AI companies encounter two distinctly new types of stakeholders and they do not pay them the attention they deserve: 👩🏻‍💻 the people whose data is used to train AI models (human data sources) 🎯 the individuals to whom AI is applied (human objects of AI) ☢ with regards to product #safety: AI models' inherent lack of transparency constitutes one of the key impediments to ensuring their safety. Hans Jonas' precautionary principles, which states: “Act so that the effects of your action are compatible with the permanence of genuine human life” or, conversely, “Do not compromise the conditions for an indefinite continuation of humanity on earth”, seems to be primarily disregarded by those at the forefront of AI development. 🎤 Concluding: yes, AI is exceptional in many ways, and influential figures cultivate this impression with bold claims, threats, and promises. But we should not lose sight of the fact that profit-seeking companies control the most influential AI applications. Regarding business ethics, these companies must be held to the same fundamental standards as any other company: to act with integrity and ensure that what they do and sell benefits humanity, rather than vice versa. Thanks The HEAD Foundation for the opportunity! #AIethics #AIsafety #corporateresponsibility #businessethics #accountability

24

11 Kommentare