Andrew Ng’s take on California's proposed law SB-1047 is spot on: “There are many things wrong with this bill, but I’d like to focus here on just one: It defines an unreasonable “hazardous capability” designation that may make builders of large AI models potentially liable if someone uses their models to do something that exceeds the bill’s definition of harm (such as causing $500 million in damage). That is practically impossible for any AI builder to ensure. If the bill is passed in its present form, it will stifle AI model builders, especially open source developers. Some AI applications, for example in healthcare, are risky. But as I wrote previously, regulators should regulate applications rather than technology. Technology refers to tools that can be applied in many ways to solve various problems.Applications are specific implementations of technologies designed to meet particular customer needs. Safety is a property of applications, not a property of technologies (or models), as Arvind Narayanan and Sayash Kapoor have pointed out. Whether a blender is a safe one can’t be determined by examining the electric motor. A similar argument holds for AI.” #publicpolicy #policy #ai #machinelearning #ml #compliance #technology #society #sb1047 #opensourcesoftware #oss
Ron F. Del Rosario’s Post
More Relevant Posts
-
If you'd like to get a closer look at what NetDocuments is doing with generative #AI, here's your chance.
🚨 Join us for an exciting webinar on August 2nd to learn how #ndMAX enables you to securely and responsibly apply cutting-edge #AI to your own documents and data in order to extract business intelligence and generate novel content. Attendees will also get to see the first solution in the #ndMAX series: PatternBuilder MAX. Use cases include: ➡️ Generate branded and formatted draft documents based on internal precedents and playbooks ➡️ Extract information from documents and save extracted data to a database ➡️ Instantly summarize documents like depositions and automatically save the summary to the correct workspace But the possibilities are endless! Learn more and save your spot 🤚 https://bit.ly/3rKtvws #ndMAX #AI #LegalAI #GenerativeAI #automation #LegalAutomation #PatternBuilder #PatternBuilderMAX #LegalTech #LegalWorkflow #LegalInnovation #DMS #ECM #AzureAI
To view or add a comment, sign in
-
What's better than AI? Don't get caught up in the Artificial Intelligence hype. REQQER has the same capabilities and uses ACTUAL software catered to your facility's needs! Give us a call today to learn more! #REQQER #maintenance #software #snf #alf #nursinghome #AI #artificialintelligence #compliance #regulatory
To view or add a comment, sign in
-
-
🌐 Did you know that recent executive action on #ArtificialIntelligence is pushing for responsible and secure use of AI software? The National Institute of Standards and Technology will develop guidelines to ensure businesses harness AI's potential while maintaining accountability. Stay ahead of the #DigitalTech game and unlock the benefits of AI for your business! 💡🚀 #Software #TechInnovation #DigitalTransformation #BusinessOperations
To view or add a comment, sign in
-
🚨 Join us for an exciting webinar on August 2nd to learn how #ndMAX enables you to securely and responsibly apply cutting-edge #AI to your own documents and data in order to extract business intelligence and generate novel content. Attendees will also get to see the first solution in the #ndMAX series: PatternBuilder MAX. Use cases include: ➡️ Generate branded and formatted draft documents based on internal precedents and playbooks ➡️ Extract information from documents and save extracted data to a database ➡️ Instantly summarize documents like depositions and automatically save the summary to the correct workspace But the possibilities are endless! Learn more and save your spot 🤚 https://bit.ly/3rKtvws #ndMAX #AI #LegalAI #GenerativeAI #automation #LegalAutomation #PatternBuilder #PatternBuilderMAX #LegalTech #LegalWorkflow #LegalInnovation #DMS #ECM #AzureAI
NetDocuments Webinar | A Look Inside ndMAX
pages.netdocuments.com
To view or add a comment, sign in
-
We've successfully wrapped up the first month of our 'Gen AI Use Case of the Week' series! In case you've missed any of the articles, here's a short recap of what we've covered so far: 📌 Small talk: What happens when 2 bots engage in a casual discussion, 20 hours is a row - https://lnkd.in/gFMcySAh 📌 Company knowledge hub: Gen AI powered knowledge management for a technology company - https://lnkd.in/gZY9GXVE 📌 AI Assisted Software Development: The future of software development - https://lnkd.in/gQNTrQV2 📌 Field Service: How to make troubleshooting work more efficient and reduce downtime of industrial machinery - https://lnkd.in/gwQercaB 📌 AI Lawyer: Legal insights powered by Generative AI - https://lnkd.in/g2tpYjTJ Stay tuned — there's plenty more ahead!
To view or add a comment, sign in
-
Software should never: - Succeed incorrectly - fail silently Building reliable deterministic software is notoriously hard. AI systems are "unreliable" by design which makes the problem notoriously harder. #AI #softwareengineering #reliability
To view or add a comment, sign in
-
@medium.zom Fine tunening LLM's such as chatgpt and Bard with Low Rank Adaptation (LoRA) Quote article "it proposes to freeze the pre-trained model weights and injects trainable rank decomposition matrices into each layer of the transformer architecture, greatly reducing the number of trainable parameters for downstream tasks." #AI #machinelearningalgorithms #machinelearning #artificialintelligence #neuralnetwork #transformers #generativeai #nvidia #databricks #oracleai #azureai #llm
What is Low-Rank-Adaptation (LoRa)
medium.com
To view or add a comment, sign in
-
SaaS Business Development, Sales and Partnerships | GTM Execution and Business Growth Leader | Channel Development | Business Messaging
Great way to contextualise this. Connectors between Software 2.0 & Software 1.0 are going to decide how quickly AI-first Software 2.0 products become usable, scalable, and match user expectations.
even in a world where developers have affordable access to very capable AI systems, I think the most challenging part of this game will still be building the integrations, infrastructure, data integrity, and essentially the "glue" between intelligence/software 2.0 <> 1.0. developing safe, compound AI systems that can help developers control behavior more tightly is exciting but also non-trivial, especially as experimental/alpha/beta products you love and are developing will soon have to scale, both in terms of user expectations and value.
To view or add a comment, sign in
-
🚀 Empower your LLMs with the precision they deserve! Introducing LLMWhisperer, your go-to technology for seamlessly preparing complex documents for LLM consumption. LLMs possess immense power, but their effectiveness hinges on the quality of input they receive. With LLMWhisperer, we bridge the gap between unstructured documents and LLM comprehension, ensuring optimal performance every time. Key Features: 📑 Process up to 100 pages daily, absolutely free! 💳 No credit card required 🔗 Three Output Modes: Layout Preserving, Text, Text-Dump Unlock Advanced Capabilities: 🔄 Layout Preserving Mode: Seamlessly maintain document layout, enabling LLMs to efficiently extract data from diverse designs, including multi-column layouts and tables. ✅ Checkbox and Radiobutton Detection: Convert these elements into text for seamless interpretation by LLMs, simplifying the handling of forms. 🔎 Auto Mode Switching: Automatically transition to OCR mode if text extraction encounters challenges with scanned documents. 💡 Auto-Compaction: Optimize token outputs, minimizing LLM costs without compromising on accuracy. Supercharge your LLM experience today with LLMWhisperer and unleash the full potential of your document processing workflows. Try it out now and witness the difference -> https://lnkd.in/gusJH5Ui #LLM #DocumentProcessing #AI #Technology #Innovation
To view or add a comment, sign in
-
-
Have you ever felt amazed and underwhelmed by the same technology? While it's impressive how much assistance we get from generative AI in optimizing and refining code, it's equally surprising to see its limitations in recalling previously given instructions. This contradiction highlights the importance of integrating higher-level functionalities into our specialized engines, significantly enhancing their usability and value. By addressing this gap, we can transform a valuable tool into an indispensable ally in our tech journey. Follow us for more updates, and be the first to know when we will release it for general use.
To view or add a comment, sign in