𝗪𝗵𝗮𝘁 𝗶𝘀 𝗠𝗶𝗰𝗿𝗼𝘀𝗼𝗳𝘁 𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗖𝗼𝗽𝗶𝗹𝗼𝘁?

"It is a generative AI-powered security solution that helps increase the efficiency and capabilities of defenders to improve security outcomes at machine speed and scale, while remaining compliant to responsible AI principles."

The primary focus of the Early Access Program is centered around:

📌𝗜𝗻𝗰𝗶𝗱𝗲𝗻𝘁 𝗿𝗲𝘀𝗽𝗼𝗻𝘀𝗲

📌𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗽𝗼𝘀𝘁𝘂𝗿𝗲 𝗺𝗮𝗻𝗮𝗴𝗲𝗺𝗲𝗻𝘁

📌𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗿𝗲𝗽𝗼𝗿𝘁𝗶𝗻𝗴

"Here's an explanation of how Microsoft Security Copilot works:

➡ User prompts from security products are sent to Security Copilot.

➡Security Copilot then pre-processes the input prompt through an approach called grounding, which improves the specificity of the prompt, to help you get answers that are relevant and actionable to your prompt. Security Copilot accesses plugins for pre-processing, then sends the modified prompt to the language model.

➡Security Copilot takes the response from the language model and post-processes it. This post-processing includes accessing plugins to gain contextualized information.

➡Security Copilot returns the response, where the user can review and assess the response."

https://learn.microsoft.com/en-us/security-copilot/microsoft-security-copilot

#microsoft #microsoftsecurity #securitycopilot #copilot #soc #incidentresponse #soc #analyst #securityanalyst #ai #artificialinteligence #generativeai #openai #azureopenai #llm #cybersecurity #defender #xdr #sentinel #intune #prompt #largelanguagemodel #llm #foundationalmodel #gpt4 #gpt3

My article on "Truthiness by Design" — building a #Legal #LLM as a #FoundationalModel, to provide:

1. Truth / Facts

2. Guardrails for autonomous agents (e.g., #AutoGPT, #BabyAGI)

https://www.linkedin.com/pulse/legal-llm-foundational-models-truthiness-design-damien-riehl

On #AI, #LLMs (e.g., #GPT-4, #Dolly2), consider:

1. #Law binds everyone (ignorance ≠ excuse)

2. Law = "free" — but in US, costs money (e.g., PACER's $0.10/page)

3. Access to Justice (#A2J) = massive problem. 80% of #legal needs unmet

RESULT: Best justice money can buy?

What if a solution — #FoundationalModel for law — helps all of the above?

RELATED: The UK is investing £100 million to build a #FoundationalModel #LLM. Government #data = LLM's foundational model

https://www.engadget.com/the-uk-is-creating-a-100-million-ai-taskforce-143507868.html

#LLM #chatbot revolution is happening and the #TechnologicalSingularity is upon us. However, the future is not distributed equally.

We are dependent on #APIs. Even with leaked weights of #LLAMA, no one would dare to run their business on those due to legal risks.

Will #Meta even want to offer their #FoundationalModel to businesses other than their own? Why would they?

With APIs, the rug can be pulled from under you and your #business for no reason at all at any time. Perhaps they want to charge more? Perhaps you compete with a service they or their #shareholders have? Perhaps they change or deprecate the API you are dependent on? Perhaps they think you violated their terms of use and you have no recourse?

Anyhow, we need #DigitalSovereignty more than ever. We can't build a world which depends on four #corporations and their graces.