DeepLearning.AI’s Post

“I’m not a man, I’m a high-performance machine.” Well, Knight Rider predicted the future: Our AI voice assistant called KITT understands you better than existing solutions. 💡 That’s because it is based on Large Language Models. 🛠 And our researchers, Raphaël Frank and Sasan Jafarnejad are working to make this happen. #ArtificialIntelligence, #AutonomousCars, #ResearchInnovation, #ResearchUnleashed

I have been learning and tinkering with AI Agents & LLMs. While trying out different tools, an example is this using AgentGPT via Reworkd (YC S23) to build an AI agent for connecting newcomer & minority founders with resources. Check this out & let me know your feedback https://lnkd.in/dPnQp_Q3

AI models that don't just chat. They get to work. Meet our new xGen-Sales and xLAM AI models that: ⚡️ Power autonomous sales tasks 🦾 Outperform models 8x their size 📈 Top the BFCL leaderboard, beating GPT-4 and Claude in function calling Learn more: https://t.co/9frYuLRGSJ https://t.co/RCzESF334D https://bit.ly/4d4Csm9

AI models that don't just chat. They get to work. Meet our new xGen-Sales and xLAM AI models that: ⚡️ Power autonomous sales tasks 🦾 Outperform models 8x their size 📈 Top the BFCL leaderboard, beating GPT-4 and Claude in function calling Learn more: https://sforce.co/3ZgliyS

AI models that don't just chat. They get to work. Meet our new xGen-Sales and xLAM AI models that: ⚡️ Power autonomous sales tasks 🦾 Outperform models 8x their size 📈 Top the BFCL leaderboard, beating GPT-4 and Claude in function calling Learn more: https://sforce.co/3ZgliyS

Since I'm working on multiple AI initiatives currently, I have a keen interest in keeping abreast of the impending regulations on the use of AI. In the absence of any federal standards or regulations passed by Congress, each state is coming up with its own legislative actions. The closest thing to federal guidelines I've seen is the following OMB memo "M-24-18 Advancing the Responsible Acquisition of Artificial Intelligence in Government": https://lnkd.in/exsKUR_z According to the National Conference of State Legislators (NCSL) there were 480 different legislative actions and/or bills that were either pending or enacted or failed etc. Here's a good summary of the 2024 AI Legislation according to ncsl.org: https://lnkd.in/erhdPKZS Finally, here's a summary of the latest California legislative activity that made the news this past week: https://lnkd.in/eJvg5Gu8

Safe ai requires data that has the appropriate governance and entitlement policies in place before analysis.

become a prompt engineer! (that might be a waste of time) TL;DR: An AI tool that generates 3D models from voice prompts with sliders for detailed adjustments, giving users more control than traditional prompt-only models. https://lnkd.in/eYq-WJQQ

Good summary on Global Regulatory Development about #AI #Risk by the The Brookings Institution - worth reading 👍 while I do not agree with everything. The US and the EU are taking divergence approaches, whilst the voluntary G7 Code of Conduct on AI meant to align the differences. Version 1.0 of National Institute of Standards and Technology (NIST) framework have much room for improvements as its focus seems to be off if compared to #Asimov robotics laws or my view: https://lnkd.in/e7crUEN8 I welcome more debates as the human race collectively would need to face the #Cyberpunk Era of Machine/ Human crossover. https://lnkd.in/eCuxWaTa

AI winter is coming, most notable points from Erik Larson: *) Self-driving cars: "if AI were headed toward human-level intelligence, self-driving cars would be making steady progress toward Level 5. Instead, the major players have largely given up. *) LLM's: "the technology is inherently “epistemically unstable,” as the generative approach undergirding it cannot tell the difference between high-probability words amounting to bullshit sentences, and truth." *) "LLMs lack an understanding of the real world and will never lead to AGI.LLMs lack an understanding of the real world and will never lead to AGI." *) "It’s about building a dynamically learning flexible intelligence. We’re not working on that."