[AI Gen] No generative AI... without data structuration
If data is the driving force behind Artificial Intelligence, today unstructured data - text, image, audio - is at the heart of the reactor.
Find out here why it's imperative for companies to invest in the collection and processing of unstructured data.
Fabrice Asvazadourian and Yves Caseau, CIO of the Michelin Group, discussed this topic in the first edition of En quête d'IA/AI Uncovered, on BFM Business.
#AI#GenAIMICHELIN IS & Digital
🚀 Dive into the Future of AI with SNCF Group GPT!
Excited to introduce SNCF Group GPT, our bespoke generative AI solution, brought to you by Laurent Hardy, Head of 574 IDF at SNCF Group Digital Tech. Experience the revolution of AI right within our networks and beyond!
🌟 Key Highlights:
💡 **Custom AI Integration**: Tailored specifically to meet our unique needs and challenges.
⏱️ **Efficiency Boost**: Streamline operations and boost productivity by automating complex tasks.
🌍 **Sustainability**: Drive environmental benefits through optimized resource management.
Curious about how we’re implementing this cutting-edge tech? Dive into our journey and discover the transformative power of AI!
👉 [Learn More](https://lnkd.in/gBSt5y9C)
#AI#Innovation#Technology#SNCF#GenerativeAI#FutureTech#DigitalTransformation#LaurentHardy#574IDF
New ExperT diaries !
🚀 Discover the Journey of AI Evolution, with Gauvain Girault, our Business Computing and AI CTO.
Our latest video explores the incredible evolution of Artificial Intelligence, starting with symbolic AI and rule engines providing fascinating results.
Gauvain Girault is approaching in this video key topics:
🔍 From Symbolic AI to Generative AI: A Fascinating Transformation
🌟 The Game-Changer: Generative AI
🔍Transparency in AI: Ensuring Trust and Safety
🔍 Trustworthy AI R&D: A Key Investment Area
👉 Watch Now and Learn More!
Next week : Light on how Eviden implements this within our new BullSequana AI offering.
#AI#GenerativeAI#DeepLearning#BigData#MachineLearning#TransparencyInAI#TrustworthyAI#Innovation#Technology#EnterpriseSolutions#BullSequana#ComputerVision
🤖 Exploring the Future of AI with Lex Fridman and Yann LeCun
If you've got a few hours to spare and are curious about the future of AI, particularly why large language models (LLMs) won't lead us to artificial general intelligence (AGI), I highly recommend tuning into a fascinating podcast featuring Lex Fridman and AI luminary Yann LeCun.
https://lnkd.in/eHHEE39x
🎧 Here’s why it’s a must-listen:🎧
Scale of Training: Modern LLMs are trained with approximately 10^13 tokens (10 with 13 zeros). Impressive, right? But consider this: by the age of four, a human has been exposed to around 10^15 data points (10 with 15 zeros!). This massive exposure means while LLMs may appear fluent, they lack a comprehensive model of the real world.
Real-World Application: Despite years of training and billions of dollars invested, tasks that are trivial for humans, like driving a car or setting a table, remain out of reach for AI. This illustrates Moravec's Paradox: AI finds tasks easy that humans consider complex (such as computation) but struggles with tasks humans find easy (such as setting the table).
Future Directions: According to LeCun, overcoming these challenges will require a new generation of AI models. These models must not only process vast amounts of data but also understand and model the physical world through observation, much like humans do. He is a proponent of the Joint Embedding Predictive Architecture (JEPA).
🔍 Key Takeaways:
Technological Demand: If LeCun's predictions hold true, future progress in AI will demand even larger and more computationally intensive models. Think bigger GPUs and increased electricity consumption.
Transformative Potential: LeCun is optimistic about AI's role in our future, comparing its potential impact to that of the printing press, which spurred the Enlightenment but also led to centuries of conflict. It's a reminder of the transformative yet disruptive power of technology.
#ArtificialIntelligence#LexFridman#YannLeCun#TechTalks#Innovation
Unleash the Power of Multimodal AI with LLMs! 💥
Discover how LLMs are breaking boundaries by processing text, images, and more! 🤯
This blog explores:
- The concept of multimodality in AI
- Key components and technologies behind multimodal LLMs
- Leading multimodal models and their applications
- Challenges and future directions
Ready to dive into the world of multimodal AI? 🚀
https://hubs.la/Q02KvQyy0#AI#LLMs#multimodal#machinelearning#technology#data#datascience#innovation
Over the past few days, we've been talking about our expert team's participation in the Digitization Specialist Meeting in Warsaw, Poland organized by Berne Union.
When discussing digitalization, AI must also be considered. In AI, balancing complexity and transparency is crucial. Tinubu Innovation LAB's latest White paper, "The Augmented Underwriter", looks at this issue and argues for AI that is not only advanced but also explainable. The debate on explainable AI (XAI) is not only technical but also ethical, as it aims to make AI accountable. How can we reconcile the sophistication of AI with the need for transparency? Let's take a look at this article by Ayoub El Qadi, Ph.D, Tinubu's Lead Data Scientist.
➡️https://hubs.ly/Q02qJ6hd0Benjamin LE FORESTIER, Marc Meyer, Olivier SAINT-RAYMOND#AI#XAI#EthicsInAI#datascience#innovation#futuretech
Consulting CIO | Intrapreneur | Mentor | Fintech , Healthtech , Cloud Security | Advisor , Architect , Analyst & Auditor
3moWould like to know , on what is happening in Sopra Steria Next Sopra Steria , understanding competition , value creation and optimized use cases .