Henri Jouhaud

‘Le Chat’ ‘la platform’ Mistral AI artificial intelligence, France’s Startup Darling, Takes Aim at the US Market : Mistral's edge could come from customers that want to avoid being locked into particular cloud companies “Customers are even saying ‘Le Chat’ and ‘la platform’ It seems that the French touch is having an effect.” opportunity is huge — Bloomberg Intelligence estimated that the software and cloud market in generative AI market could reach $350 billion by 2032 : Mistral is going after OpenAI’s home turf “They do not want to be forced into an AI stack” “They want choices” Mistral released a series of AI models, which power generative AI services, and a ChatGPT-like feature called Le Chat : US manager says businesses are looking for alternatives to Big Tech : will find customers through these partners as well as via direct deals with companies using its open-source models. Open source is one of Mistral’s main selling points : the approach — where the underlying computer code is openly shared and tweakable — as a safer, more versatile method than closed systems offered by OpenAI and others : Mistral’s open-source models on-premises, with data stored in local servers instead of with public cloud providers ---- generative artificial intelligence ---

1

1 commentaire

Today is the first OpenAI DevDay for this autumn. No new models are expected, but OpenAI has promised to introduce some long waited features to APIs. This is the schedule (in San Francisco timezone): - 10:00 - Welcome + kickoff by Olivier Godement & Romain Huet - 11:00-11:45 - Topic to be revealed during keynote - 12:00-12:45 - Topic to be revealed during keynote - 12:45-12:00 - Lunch - 2:00-2:45 - Topic to be revealed during keynote - 3:00-3:45 - Topic to be revealed during keynote - 4:00-4:45 - Fireside chat with OpenAI leaders - 4:45-7:00 - Closing reception So reading this, there could be up to four new announcements coming. Here is my wishlist in priority order: - Audio/voice multimodal API for gpt-4o (so advanced voice mode API) - Function call webhooks, so that AI could call app function in mids of voice-call or o1 reasoning sessions. - New gpt-4o image generation API - Prompt caching - Web search enabled api (ala Perplexity or Command-r) But I hope they leave the best for London in end of month, where I'm going 😂

12

1 commentaire

Deep tech innovation: Your key to differentiation in the new NTN economy   🛰 As the Non-Terrestrial Network (NTN) ecosystem rapidly evolves, staying ahead of the competition is no longer just about being present—it's about differentiating your offering. So how can you stand out in this crowded, fast-moving market? Our latest white paper ‘𝗔𝘀𝘀𝗲𝗺𝗯𝗹𝗶𝗻𝗴 𝗿𝗲𝗯𝗲𝗹 𝗮𝗹𝗹𝗶𝗮𝗻𝗰𝗲𝘀 𝘁𝗼 𝗰𝗼𝗺𝗽𝗲𝘁𝗲 𝗶𝗻 𝘁𝗵𝗲 𝗻𝗲𝘄 𝗡𝗧𝗡 𝗲𝗰𝗼𝗻𝗼𝗺𝘆' is now available to download https://hubs.li/Q02PhWPd0   At CC, we believe deep tech innovation holds the answer. From free space optics to AI, our whitepaper dives into how harnessing radical science and engineering can give you an unassailable commercial edge.   Key deep tech areas explored: Free Space Optics (FSO) for high-speed, low-latency data transmission. AI for autonomous satellite networks and dynamic optimisation. Advanced antennas and beamforming for superior signal strength and coverage and more...   The future of NTNs is happening now, and deep tech is your gateway to market leadership. Discover how to transform your service offerings and unlock your unique competitive advantage.   #WSBW #DeepTech #SatelliteIndustry #NTN #Innovation #GlobalConnectivity #Satellite #Connectivity

8

Your contribution for the evening 🌖: OpenReplay👇 Led by Mehdi Osman, OpenReplay is an elegant open-source session replay stack that lets you see what users do on your web app, helping you troubleshoot issues faster. Here’s why you should contribute to this repo: 1️⃣ OpenReplay offers an advanced developer tool suite that facilitates session replay, error tracking, and real-time user interaction analysis. With around 90 issues currently open for grabs, contributing allows you to improve how developers diagnose and resolve issues, ultimately enhancing user experience on web and mobile applications. 🖥️ 2️⃣ If you like a challenge, you are in luck! At Quira, we rank Developers by their DevRank (more on this here: https://lnkd.in/eNtrs97F). We found that the median developer working on the repo belongs to the ~97th percentile of GitHub developers. If you want to challenge yourself and contribute to a growing repo with over 8,000 stars, here's your shot! 🏀 3️⃣ Maintainers take extra care in getting back to you with an 85% reply rate and an average of 2 days to answer your queries. This means your contributions will get the attention they deserve. 🌟 You can check the GitHub page directly here: https://lnkd.in/dEM3SQh 📖

3

The replay of our last edition at Doctolib is here ! We talked about to talk about “Organizing data teams for AI success” in collaboration with Doctolib. ********* Program ********* 1️⃣ How can data teams serve a bold AI vision? Anouk Barnoud, Senior Data Scientist Doctolib Diana Carrondo, Data Governance Lead Doctolib Lucille FARGEAU, Senior Analytics Engineer Doctolib 💥 Let’s deep dive into Doctolib's vision of how Artificial Intelligence can reinforce our positive impact on the millions of patients and healthcare professionals we serve in Europe. This is actually the story of how much we care about our data value chain, facing challenges and learnings like in any adventure. 2️⃣ Malt's journey in building an Internal Generative AI product for Efficiency Anaïs Ghelfi, Head of Data Platform Malt 💥 Discover how Malt designed and implemented an internal Generative AI product to enhance the team’s efficiency. Anaïs will discuss the project's journey from initial concept to final implementation, including the organizational structure, onboarding, adoption challenges, and technical hurdles faced along the way. 3️⃣ Strengthen Your Core: How a Robust Core Data Product Foundation Fuels Analytics and AI Capabilities Gaëlle SERET, Group Product Manager & Staff, Core Data Product team Decathlon Digital Enjoy !

39

2 commentaires

Explainable AI The use of #AI models is becoming increasingly evident. Even if some of them are integrated into black-box decision-making systems. The explinability encourages the acceptability by users. They give more confidence in the produced decisions. In finance, healthcare and education, we simply CAN'T accept black-box AI result.  Several packages are available in #Python to produce the explanibility : ✅ #SHAP (SHapley Additive exPlanations) is inspired by the game theory. It determines how the “players” share the winnings. The sum of the contributions of each player is equal to the total gain. SHAP estimates the contribution (share) of each model input in the produced result and proposes representations of this distribution It’s agnostic to selected model. ✅ #LIME ( Local Interpretable Model-agnostic Explanations ) explains the prediction of any classifier in an interpretable and faithful manner by learning an interpretable model locally around the prediction. It’s a model agnostic. ✅ AI Fairness 360 toolkit is an open-source library containing techniques to help detect and mitigate bias in machine learning models. Package contains a set of metrics for datasets and models to test for biases, their explanation and suggestions how to mitigate. ✅ #PyXAI (Python eXplainable AI) is a library that allows to bring formal explanations suited to (regression or classification) tree-based ML models.  They claim always generate always correct explanations. Is explainability a pain for your project ? What package do you use to get it? 

36

9 commentaires

Our CMO and VP Ecosystem Isabelle Turcotte answered journalist Bruno Guglielminetti's questions live from Europe's biggest technology event, Viva Technology, last week in #Paris. She underlined the steady progression of the Canadian presence and spoke of the participation of numerous Canadian companies, illustrating Canada's ability to position itself favorably for the future of artificial intelligence. "For Canada, this is a super-important issue for productivity. If we really want to be able to improve our productivity, and grow economically, we have to make sure that our companies transform their processes or improve the way they do things, and AI is a good solution for that." 💬 Listen to the podcast here 👉 https://lnkd.in/eUS2N4Ub #AI #ArtificialIntelligence #Technology #Startup

5