185 | 🙅👨🏾‍⚖️ Nope. No, to AI-enhanced evidence.

In today’s Brainyacts:

1. Enhancing evidence with AI

2. Interviewing your personal AI

3. A simple breakdown of OpenAI ownership and other AI model news

4. US Circuit Court Judge states banning lawyers from AI is misplaced and other AI-related content

👋 to new subscribers!

To read previous editions, click here.

Lead Memo

🌊 🚨 A Watershed Moment: AI's Role in the Courtroom

I recently noted this case in the news section but I want to do a deeper dive and share it with you.

In a potentially groundbreaking decision in Washington state, a judge banned AI-enhanced video evidence in a triple murder case. This ruling marks a critical juncture in the legal world, distinguishing between the use of artificial intelligence as a tool for evidence enhancement and the unacknowledged creation of evidence, like deepfakes.

Superior Court Judge Leroy McCullogh's decision in King County reflects a growing unease with AI's role in legal proceedings. He expressed concerns that AI-enhanced evidence could complicate and prolong trials, as it involves non-transparent processes not subject to peer review. This stance was taken in the case of Joshua Puloka, accused of murder following an incident captured on smartphone video. Defense efforts to use AI to clarify this video were rejected because the enhancements were deemed potentially misleading, substituting AI’s interpretations for actual evidence.

The defense's intent was to make the video evidence clearer using technology from Topaz Labs, which promises to enhance video quality. If you go to their website’s homepage, you will see a demo of what they can do. It’s a picture of a tiger. Does this show too much AI enhancement so as to distort the truth?

Well according to the judge, this action raised significant questions about the reliability of such enhancements. The prosecution and the judge highlighted that the AI methods could introduce inaccuracies, presenting a version of events shaped by the AI's algorithms rather than the factual record.

This skepticism towards AI-enhanced evidence in courtrooms echoes broader concerns about deepfake technologies. These AI-generated images and videos can create convincingly false representations of reality, challenging the integrity of evidence. While the current case involves known and disclosed AI usage, it signals the judiciary's cautious stance on any form of AI-manipulated evidence, whether openly used or surreptitiously created as in the case of deepfakes.

The FBI's forensic analysis, indicating that the enhanced video showed elements not present in the original, underscores the dilemma. While AI enhancements might look more clear or appealing, they can distort the truth, presenting a fabricated clarity. This discrepancy between perceived clarity and factual accuracy is at the heart of the court's decision.

This ruling could serve as a precedent, discouraging legal reliance on AI for evidence enhancement due to the risk of misrepresenting the truth. It highlights the need for a more nuanced understanding and regulation of AI technologies in legal settings, differentiating between legitimate technological aids and manipulations that could jeopardize justice.

Spotlight

🗣️🤖 Future Job Interviews

Short Spotlight section today as I am onsite for a client’s retreat. But I wanted to share this intriguing and mindblowing thought.

We had Mustafa Sulleyman join us for a fireside chat. Mustafa is now the CEO of Microsoft AI and the co-founder of both DeepMind (purchased by Google) and Inflection AI (makers of chatbot Pi and soon going into the commercial space).

Mustafa suggested that soon we will all have personal AI bots/agents. They will be instrumental in getting our work done, especially in knowledge-intensive domains such as legal.

Because of this, when interviewing for a job or pitching to a client, it will become commonplace to “interview” the personal AI bots/agents too. What this will look like? Who knows. But it raises an interesting question - if we each have an AI bot help us, ought we not be prepared to open them up to some sort of evaluation for hiring? If the employer/client is likely to interact directly with it or its work product - it makes a ton of sense. But the thought of this makes my head hurt

AI Model Notables

► AI Laptop Wars: Microsoft is confident its forthcoming AI PCs could finally beat Apple.

► Apple’s $25-50 million Shutterstock deal highlights fierce competition for AI training data.

► AI hardware company from Jony Ive (former Apple chief designer) and Sam Altman (OpenAI CEO) seeks $1 billion in funding.

► Tesla’s Robotaxi coming in August (we think) and here is what that might look like.

► A great breakdown of who owns and influences OpenAI.

► The NY Times suggests that GPT-4 was trained on transcripts from one million hours worth of YouTube videos with or without Google directly involved.

► Google App for Android could get a Gemini AI toggle switch plus Gmail app to get email summary feature.

► Meta's approach to labeling AI-Generated content and manipulated media.

► Spotify Premium users can now turn any idea into a personalized playlist with AI playlist in beta.

News You Can Use:

➭ US lawmakers unveil sprawling plan to expand online privacy protections.

➭ Blind internet users struggle with error-prone AI aids.

➭ New York lawyers urged to use AI with care in new state bar guidelines.

➭ Trudeau unveils $1.8 Billion package for Canada’s AI sector.

➭ U.S. Circuit Judge John Nalbandian argued that bans on lawyers using AI programs like ChatGPT to conduct research were "misplaced" efforts that would also bar potentially beneficial uses of the technology by litigants with limited means.

➭ AI-animated Pink Floyd music video stirs debate.

➭ JPMorgan Chase CEO Jamie Dimon said the firm has already grown its AI talent to over 2,000 experts and data scientists, with over 400 AI/ML use cases in production. 

➭ Joint Statement EU-US Trade and Technology Council of 4-5 April 2024 in Leuven, Belgium.

➭ Union Minister Ashwini Vaishnaw says India to legislate AI regulation soon.

Who is the author, Josh Kubicki?

Some of you know me. Others do not. Here is a short intro. I am a lawyer, entrepreneur, and teacher. I have transformed legal practices and built multi-million dollar businesses. Not a theorist, I am an applied researcher and former Chief Strategy Officer, recognized by Fast Company and Bloomberg Law for my unique work. Through this newsletter, I offer you pragmatic insights into leveraging AI to inform and improve your daily life in legal services.