Benjamin Wolba

ETH Zurich AI have pulled theoretically validated an auction mechanism that aggregates multiple agents' preferences over LLM-generated replies without the need for fine-tuning. The work primarily focused on the multi-agent preference optimization in online advertising, can adeptly balancing advertiser interests with the integrity of user-focused content. The researchers address the significant challenge of preference aggregation where multiple stakeholders have divergent expectations from LLM responses. Traditional methods, which might blend these preferences, often result in diluted or incoherent outputs that satisfy no one. The novel mechanism introduced by the team uses auctions to navigate this complexity, ensuring that LLM outputs align more closely with the aggregated preferences of all involved parties. A key innovation of their approach is the use of a non-fine-tuning auction mechanism that converges toward the performance of an optimally fine-tuned model as computational resources increase. This mechanism, requiring only API access to LLMs, sidesteps the logistical and ethical issues associated with direct access to model weights, fostering a more flexible and scalable integration into existing systems.   Quite cleverly, the mechanism's design incorporates features that encourage truthful reporting of preferences by agents. It introduces a payment rule that aligns each participant's utility with their contribution to the overall social welfare. This feature is crucial in maintaining the system's integrity and long-term viability, as it ensures that participants are incentivized to contribute positively to the collective output.   Experimental results underscore the mechanism's efficiency and its ability to enhance both advertiser value and platform revenue significantly. Advertisers can influence LLM-generated responses to subtly promote their products while maintaining content relevance and user satisfaction. This balance is vital for platforms seeking to maximize advertising revenues without compromising user experience. The research not only contributes a novel mechanism to the field of auction design but also extends its applicability to the management of LLM outputs in commercial settings. Arxiv: https://lnkd.in/eTSR6ZHg

1

📢 Exciting News from ETH Zurich! We had the pleasure of welcoming Gavin Wood, one of the most influential figures in the blockchain space, to ETH Zurich. Gavin shared insights on the JAM Protocol, a groundbreaking innovation that promises to redefine scalability in decentralized systems. 🚀 JAM Protocol Highlights: Novel Non-Synchronous Approach: Unlike traditional protocols that rely on synchronous operations, JAM Protocol introduces a non-synchronous mechanism, enhancing scalability without compromising security. Enhanced Efficiency: By decoupling operations, JAM Protocol significantly reduces bottlenecks, allowing for faster transaction processing and improved network throughput. Robust Security: Despite its innovative scaling techniques, JAM Protocol maintains rigorous security standards, ensuring reliable and secure operations. Versatility: This protocol can be seamlessly integrated into various blockchain ecosystems, providing scalable solutions tailored to diverse use cases. 🌟 Why It Matters: Scalability: Addresses one of the most pressing issues in blockchain technology, enabling networks to handle increased demand. Innovation: Pioneers a new paradigm in blockchain protocol design, setting the stage for future advancements. Thank you, Gavin Wood, for sharing this transformative vision with us. The future of blockchain scalability looks brighter with innovations like JAM Protocol leading the way. #Blockchain #Scalability #Innovation #ETHZurich #JAMProtocol #GavinWood

65

2 Kommentare

According to Odaily, a study conducted by consulting firm Universelle has found that the #SRC20 protocol surpasses other protocols running on the #Bitcoin network due to its 'unmatched data persistence'. The upcoming cost reduction upgrade of the protocol is set to make SRC20 one of the choices for long-term projects and investments. The study suggests that the SRC20 protocol is superior to #Runes and #BRC20 as it does not require the #transmission of inscriptions or splitting of #UTXO. In terms of wallet support, the report notes that while BRC20 and Runes have a higher level of infrastructure and wallet support, the SRC20 protocol is rapidly catching up.📊 In terms of funding, projects based on SRC20 have attracted venture capital ranging from $5 million to $10 million, which is less than BRC20 ($20 million). However, Universelle asserts that the capital flowing into #SRC20 tokens will increase as acceptance and adoption levels rise. #tokens #cryptocurrency #binance #coins #tokonomics #finance #web3 #investments #funding

20

$1.3B in 2024 and expected to be $5.3 billion by 2029; the Quantum market is growing at a CAGR of 32.7 (MarketsandMarkets™) One quantum technology that is more mature than any other and has far-reaching potential is “𝙌𝙪𝙖𝙣𝙩𝙪𝙢 𝙎𝙚𝙣𝙨𝙞𝙣𝙜.” Quantum sensing technology improves the accuracy of how we measure, navigate, study, and interact with the world by sensing changes in motion, electric and magnetic fields. The global quantum sensor market size surpassed $149M in 2023 and is expected to hit $1.1B by 2033. Here, we will discuss short and long-term potential use cases and some resources you might be interested in. 𝗦𝗵𝗼𝗿𝘁 𝘁𝗲𝗿𝗺 → Inspection and design improvement of batteries → Navigation based on electromagnetic fields → Cardiac diagnostic with biomagnetic → Bioimaging with enhanced MRI and PET → Targeted photodynamic therapy for cancer → Monitoring stability and integrity of infrastructures → Flaw detection of microelectronics 𝗟𝗼𝗻𝗴-𝘁𝗲𝗿𝗺 → Efficiency boost for energy conversion in solar cells → Magnetometer for fusion reactors → Determination of protein structures under physiological conditions → On-chip monitoring of security and malfunction of chips → Microscopy for verification of high-value artifacts → Mind-machine interface → Metabolic measurements in living cells Quantum sensing is poised to revolutionize many industries, but current device cost and size present obstacles that need to be addressed strategically. 𝙏𝙤 𝙠𝙚𝙚𝙥 𝙮𝙤𝙪𝙧𝙨𝙚𝙡𝙛 𝙪𝙥𝙙𝙖𝙩𝙚𝙙 𝙬𝙞𝙩𝙝 𝙦𝙪𝙖𝙣𝙩𝙪𝙢 𝙨𝙚𝙣𝙨𝙞𝙣𝙜, 𝙝𝙚𝙧𝙚 𝙖𝙧𝙚 𝙨𝙤𝙢𝙚 𝙪𝙨𝙚𝙛𝙪𝙡 𝙧𝙚𝙨𝙤𝙪𝙧𝙘𝙚𝙨: 1- IBM Quantum roadmap https://lnkd.in/dWZAzhCJ 2- Revolutionary quantum startup in Europe https://lnkd.in/dhFntg7Q 3- Quantum economy blueprint https://lnkd.in/dvtVucEB 4- QTech 2024 | Sep 10-12 | Berlin https://lnkd.in/d--aMT76 Lastly, proud to share that The Deep Tech Lab - Quantum of the BioInnovation Institute (BII) in Copenhagen has nominated APEX Ventures into their Advisory Group. As big believers and supporters of the Danish quantum ecosystem, we will help the lab select and evaluate candidates for the program and bring our expertise in the quantum sectors and company building in general. We are excited to work with Jens Nielsen, Cathal J. Mahon, Jakob Svagin, Raheesty Nishta Nem, and their team to accelerate quantum innovation in Denmark. Credit: McKinsey & Company #AI #Startups Follow us for strategies and resources for Deep Tech founders and VCs! And get access to exclusive content on deep tech startups like planqc, kiutra, and Quibim in our newsletter: 🗻 https://lnkd.in/dtTX49AF

84

3 Kommentare

🌟 Excited to speak today at the Crypto Valley Association's Sustainability Working Group Monthly Meetup! I'll be discussing 'Empowering the Grid: Blockchain Innovations in Decentralised Energy', sharing insights into how Powerledger's cutting-edge peer-to-peer energy trading is revolutionizing decentralized energy markets. Join us to explore the significant strides we're making with blockchain, AI, and other emerging technologies to create sustainable, efficient energy systems. Don't miss out on learning about our challenges, opportunities, and visionary plans! 📅 Today, 8 May 2024 | ⏱ 17:00 CET #EmergingTech #DePIN #BlockchainInEnergy #LiveSession #EnergyMarkets Looking forward to seeing many of you there!

4

In the evolving landscape of AI technology, choosing between open-source and proprietary models is more than just a technical decision, its a strategic one. As industries and developers weigh the benefits of accessibility and innovation against security and performance, the future of AI might just lie in a hybrid model that leverages the strengths of both.

16

1 Kommentar

The chips from #Nvidia that perform #ArtificialIntelligence calculations are so #valuable that they are delivered in #ArmoredCars https://lnkd.in/eMctJnCQ #fintech #chipmaker #AI #MachineLearning #BigData #GenerativeAI @asafitch @WSJ

16

OpenAI are holding a product announcement at 10:00 PST / 18:00 BST / 19:00 CET today: https://www.openai.com/ Is GPT4 getting upgraded today? Is it getting sibling products and services? CEO Sam Altman has been talking up a storm in recent weeks about the future of AI ('Universal Basic Compute' access for all global citizens? A cognitive version of 'let them eat cake' I fear). At the same time he has stated that today will not be GPT5, nor search, but will feel 'magical'. Hmmm. The only way to find out will be to listen in or come back here afterwards to discuss what they are releasing. My hope is that they build on the practical benefits of custom GPTs and add Agents, and improve the voice interface. Having well integrated search functions to allow for clear referencing would also be neat, especially if it is done on the custom knowledge that is uploaded. Maybe we'll hear more about the rumoured integration into iOS 18. Whatever it is, you can be sure that here at Curistica we'll be first to explore the strengths, weaknesses, risks, and real world use cases for you! What are you hoping for?

23

9 Kommentare

Of the top 50 Web3 protocols in 2016, ~70% have lost relevance. Read our thoughts on this below. And stay tuned for our framework of 15 core elements that Web3 leaders must consider for their protocol’s success. #DigitalAssets #Web3

3

💬 ➡️ 📽️ 𝗠𝗼𝗿𝗲 𝘁𝗲𝘅𝘁 𝘁𝗼 𝘃𝗶𝗱𝗲𝗼 𝗴𝗼𝗼𝗱𝗻𝗲𝘀𝘀 𝘁𝗵𝗶𝘀 𝘁𝗶𝗺𝗲 𝗳𝗿𝗼𝗺 Higgsfield AI Following the mind-blowing demos we have seen from SORA AI, Higgsfield AI is a similar video AI company that wants to "democratise social video creation to everyone" Their differentiation appears to be: - Unparalleled levels of camera and motion control - Realistic human characters 🧑🏽‍🦱 They state their game changing foundational model excels at creating personalized characters with lifelike motion - with just 1 selfie and all on mobile. 📱 It is not fully released yet although the underlying technology is being used in a mobile app called Diffuse which is only available in a few countries. #GenAI #TextToVideo #AI #Sora You can join their waitlist on their website: https://higgsfield.ai/ ❓ What do you think of this technology and have you tried any of the Text-To-Video tools out there? Drop me your thoughts in the comments below 👇🏽

5

15 #Graphs That Explain the State of #AI in 2024 https://lnkd.in/ecKsa3EE #fintech #ArtificialIntelligence #MachineLearning #GenerativeAI #GenAI IEEE Spectrum

17

1 Kommentar

🚨 - Alright, let's go! Not in order, but the following episodes definitely are the ones that you should not miss: Bitcoin im Bundestag 🎙️ Brace yourself for an enlightening journey into the realm of Bitcoin, right from the heart of the German Bundestag! 🇩🇪 In this exclusive episode, we delve deep with Joana Cotar, Stephan Schmidt, and Lina Seiche, unraveling their perspectives on integrating Bitcoin into German politics and society. Meet Joana Cotar, a trailblazing German politician and entrepreneur, revered for her staunch advocacy of technological innovation and decentralization. As a prominent member of the German Federal Parliament, she spearheads the drive for Bitcoin and Timechain adoption, viewing them as potent tools to fortify the economy and safeguard citizens' financial autonomy. Stephan Schmidt, an esteemed economist and financial advisor, brings his extensive banking sector experience to the table. With keen insights into financial markets, Stephan sheds light on how Bitcoin could reshape monetary policies and bolster economic stability across Germany and the European Union. Then, there's Lina Seiche, the visionary behind The Little HODLer, a celebrated Bitcoin merchandise brand adored by millions globally. Since 2021, The Little HODLer has captivated Bitcoin aficionados worldwide with its captivating comic series, empowering HODLers from over 60 nations through plushies, prints, comic books, and more. Here's a sneak peek into the show's highlights: ✅ Integration of Bitcoin into German politics: Explore how policymakers in the Bundestag are navigating Bitcoin regulation and its potential ramifications for the national economy. ✅ Economic perspectives: Joana, Stephan, and Lina unravel the implications of Bitcoin on the traditional financial ecosystem, spotlighting opportunities and challenges for investors and consumers alike. ✅ Social impact: Delve into Bitcoin's transformative potential in uplifting marginalized communities and fostering financial inclusivity across Germany and beyond. Tune in for a riveting discussion that promises to redefine your understanding of Bitcoin's role in shaping the future of Germany's political and socioeconomic landscape. Don't miss out on this enlightening discourse! 🚀 #Bitcoin #Germany #Bundestag https://lnkd.in/dJhg_tNF

1

NVIDIA gives developers a free, scalable way to generate synthetic data that can be used for training and refining LLMs. This is particularly useful to optimize for domain specific tasks (e.g. #legaltech).. But how will this impact synthetic data generation startups? #ai #artificialintelligence #ArtificialIntelligence #MachineLearning #DeepLearning #SyntheticData #DataScience #LLM #NLP #OpenSource #AIResearch #TechInnovation #DataGeneration #AIModels #HuggingFace

9

3 Kommentare

#AGoodRead #Senate unveils $32 billion #roadmap for regulating AI #Calling for at least $32 billion to be spent each year for #nondefense #AI #innovation. #Appendix Consists of very interesting AI Insight #Forum #Participants https://lnkd.in/dw2mekfH

3

New spin-off from the SRI / NSG labs at ETH Zürich, solving the network observability problem with GenAI & mathematical models, based on decades of research work on automated reasoning for network verification and synthesis. Follow NetFabric.ai to track where this goes :)

31