Writing assistance, code generation, and conclusions over data - How machine learning and artificial intelligence generate and understand natural language.
The whitepaper by Dr. Michel Rath and Nicki Borell explains the current state of technology, what OpenAI and Microsoft are doing, and how interested customers can benefit from it.
The first part covers the architecture and technical details of Generative Pre-trained Transformers, or GPT for short. It covers basic concepts such as LLM's - (Large Language Model) and what a Prompt is. The whitepaper explains the difference between OpenAI, Azure OpenAI and the announced Microsoft Copilot feature.
Details:
- Introducing ChatGPT
- Basic terms
- LLM
- Prompt
- GPT3, GPT4 and other models
- The cooperation between Open AI and Microsoft
- Microsoft Copilot
- What is the difference between Microsoft Copilot and Microsoft OpenAI
- Current state of technology - what is available, what is announced?
Details:
- Legal Aspects related to ChatGPT & Co.
- ChatGPT and data protection
- Guidelines for employees and the protection of trade secrets
- ChatGPT and Copyright
- The AI Regulation is coming
Download
- Download German version: LINK
- Download English version: LINK
Authors
Dr.Michael Rath is a
lawyer, a specialist in information technology law and a partner at Luther
Rechtsanwaltsgesellschaft mbH, based in Cologne. He coordinates Luther's
Information Tech & Ccommunications practice area. He is also a Certified
ISO/IEC 27001 Lead Auditor.
Nicki Borell is co-founder of Experts Inside, a technology consultancy focused on Azure and Microsoft 365, and the head behind the label "Xperts At Work". His focus topics are enterprise collaboration, security and compliance. The Azure OpenAI services and the GPT language model therefore fit perfectly into his work context. They build another integartion between Microsoft 365 and Microsoft's Azure services. Content generation and semantic search are thus also possible for the data within a Microsoft 365 environment in a secure and controllable way.
Keine Kommentare:
Kommentar veröffentlichen