OpenAI’s head scientist, Ilya Sutskever, revealed in an interview by a fellow scientist near his level, Sven Strohband, how the emergent intelligence of his neural-net AI was the surprising result of scaling, a drastic...more
In a landmark move towards the regulation of generative AI technologies, the White House brokered eight “commitments” with industry giants Amazon, Anthropic, Google, Inflection, Meta, Microsoft, and OpenAI. The discussions,...more
8/1/2023
/ Artificial Intelligence ,
Cybersecurity ,
Data Security ,
Discovery ,
e-Discovery Professionals ,
Electronically Stored Information ,
Machine Learning ,
Regulatory Oversight ,
Risk Management ,
Technology Sector ,
Technology-Assisted Review
This article continues the Ai creativity series and examines current thinking among lawyers about their work and job security. Most believe their work is too creative to be replaced by machines....more
This article analyzes the creative imagination capacities of ChatGPT and two software settings to control it: Temperature and Nucleus Sampling (aka Top-P). It is easy to change these parameters, as will be shown by multiple...more
Sam Altman had some fascinating things to say recently about technology revolutions and what comes after AI. Apparently that is his favorite unasked question. After completing a grueling world tour of meetings and interviews,...more
The report by well-known consulting firm McKinsey, The economic potential of generative AI: The next productivity frontier (June 2023), provides reliable information and analysis on the jobs potential of ChatGPT and other...more
Are Both Human and AI Generated Thoughts and Writings Mere Predictable Repetitions of Prior Language? Introduction My experiences with Large Language Model AI prompt me to wonder about language, ideas and intelligence. These...more
The GPTJUDGE: Justice in a Generative AI World article will be published in October by Duke Law & Technology Review. The authors are Maura Grossman, Paul Grimm, Daniel Brown and Molly Xu. In addition to suggesting a legal...more
Help! My AI needs a psychiatrist. Either that or Open AI has some explaining to do. [Editor’s Note: This article was first published May 31, 2023 on e-Discovery Team. EDRM is grateful to Ralph Losey, EDRM Advisory Council...more
A version of OpenAI’s ChatGPT-4 is now available for iPhones and can be found at the Apple App Store. This is OpenAI’s only official iPhone app and runs both version 3.5 and the far better version, 4.0. It does not have the...more
On May 12, 2023, Chat GPT-4 began its release to the public of two new features in GPT-4, Web Browsing and Plugins. Just a few days later, a red-team instructor, Johann Rehberger, claimed to have made a successful indirect...more
1. SUMMARY OF INSIGHTS SO FAR. A scientist and tech lawyer impressions of ChatGPT, with emphasis on how it may impact the law going forward.
a. RL - "These are ALMOST the Droids we’ve been looking for!"
(i) Impressed by...more
5/22/2023
/ Artificial Intelligence ,
Bots ,
Data Collection ,
Data Preservation ,
Discovery ,
Duty of Competence ,
Duty of Confidentiality ,
Duty of Fairness ,
e-Discovery Professionals ,
Electronically Stored Information ,
Evidence ,
Fiduciary Duty ,
Information Governance ,
Law Firm Associates ,
Law Firm Partners ,
Legal Ethics ,
Legal Technology ,
Popular ,
Webinars
My last blog, ‘A Discussion of Some of the Ethical Constraints Built Into ChatGPT‘ concluded with my encouraging Red Team testing. We need hackers to prod, con, trick and manipulate Ai chatbots; to jailbreak them. We need...more
Open AI’s recent GPT-4 technical report, Open AI (2023), is a must read for all serious students of Ai. One of the most interesting parts of the report is its discussion of the efforts to build in protective ethics that are...more
There are severe, inherent limitations to all LLM types of Artificial Intelligence because their “apparent” intelligence is based solely on Language and Images. They have no real intelligence or understanding of any of the...more
Ironically, LARGE Language Models (LLMs) Have Only SMALL Language Memories; Way Too Little for Case Analysis. After just a 12,288 word input, about 40 pages of double-spaced text (25 words per page), ChatGPT-4, which is able...more
Although still at sea on an Atlantic crossing, I remain obsessed with the power of this new Ai tool. In the last blog we prompted a discussion by Chat-GPT with itself on the subject of AI and the future of the Law. Here we...more
While crossing the Atlantic Ralph Losey again uses the ChatGPT-4 writing assistant and evolving “Hive Mind” prompt. The intriguing prompt was developed at the Open AI “Prompt Engineering” channel on Discord with some special...more
I prompted ChatGPT-4 to generate a dialogue with itself through various personalities of its own choosing. I did virtually nothing after setting up the lengthy prompt. ChatGPT-4 came up with the different experts that it...more
Written by Ralph Losey with some help from ChatGPT-4. All illustrations by Losey using Midjourney or Dall-E. This is the conclusion to this blog article, which I split into two parts for ease of human reading. Please read...more
Written by Ralph Losey with some help from ChatGPT-4. All illustrations by Losey using Midjourney or Dall-E. Predicting the likelihood of the Singular We need to intensify our preparation efforts, starting now. In this...more
Introduction - Keep your Bots on a short leash for now. Image: Losey Dear Bar colleagues and fellow tech enthusiasts, this article explores the top-ten most obvious ways that Large Language Models (LLMs) such as ChatGPT-4...more
Social Progression in the Age of Personal Computers, a Necessity for Survival. Can Ai save the day? Please read Part 1 of this article first. Explained by ChatGPT, from it’s perspective, with prompt engineering, quality...more
Social Progression in the Age of Personal Computers, a Necessity for Survival. Can Ai save the day? Explained by ChatGPT, from it’s perspective, with prompt engineering, quality control and some minor contributions by...more
Prompt Engineered, Edited and Quality Controlled by Ralph Losey. This essay was co-written by a human and machine with over 80% of the text “generated” by ChatGPT. Active machine learning is a type of Artificial Intelligence...more