• The Brainyacts
  • Posts
  • 197 | šŸ—ļøšŸ‘“šŸ½ Building ChatGPT-based new lawyer mentors

197 | šŸ—ļøšŸ‘“šŸ½ Building ChatGPT-based new lawyer mentors

Brainyacts #197

Itā€™s Tuesday. There is so much drama at OpenAI right now. I am choosing not to cover most of it as it is neither pragmatic nor all that useful. If any of it represents real impact to how we use their tools, I will certainly highlight it as I have. But geez, drama drama drama!

Letā€™s get to it.

In todayā€™s Brainyacts:

  1. Lawyer mentoring via ChatGPT (as a start)

  2. Microsoftā€™s total ā€œRecallā€

  3. Immigration firm giving free AI access to clinics and other AI model news

  4. Scarlett Johansson is pissed and other AI-related content

šŸ‘‹ to new subscribers!

To read previous editions, click here.

Lead Memo

šŸ—ļøšŸ‘“šŸ½ Building ChatGPT-based new lawyer mentors

Earlier today I gave a talk to an international law firm on the impact of generative AI on legal education and lawyer development. One of the critical areas impacted by GenAI is the training and development of new lawyers. With traditional methods of mentorship facing challenges in scalability and consistency, prototyping a mentor using ChatGPT offers a promising alternative.

A typical week in a new lawyerā€™s learning life (pre-2020)

A typical week in a new lawyerā€™s learning life in 2024

The Necessity of AI Mentorship

  1. Shrinking Learning Opportunities: The surface area for new lawyers to learn through traditional on-the-job experiences is shrinking. Generative AI tools are increasingly handling routine tasks that used to be learning opportunities for new lawyers, such as contract reviews and memo writing. This shift means new lawyers have fewer chances to develop practical skills through direct work experiences.

  2. Hybrid Work Environment: The shift to hybrid work models has reduced the frequency of serendipitous, ad hoc learning interactions that naturally occur in an office setting. With fewer opportunities for spontaneous mentoring, a structured and scalable solution is needed to ensure consistent learning and development.

  3. Client-Driven Changes: Clients are beginning to use generative AI tools themselves, further reducing the routine tasks they delegate to law firms. This self-sufficiency by clients reduces the volume of work available for training new lawyers, exacerbating the need for alternative training methods.

  4. Diverse Learning Styles: Modern educational theories recognize that individuals have diverse learning styles. Personalized learning journeys, which can be facilitated by AI, are becoming crucial for effective education and professional development. This personalization is challenging to achieve in traditional mentorship programs but can be efficiently addressed with AIā€‹ā€‹

Hereā€™s how law firms can leverage ChatGPT to prototype a mentor for new lawyers, addressing the necessity and feasibility of this approach.

Follow along with this brief chat interaction that took me less than 2 minutes of my time to create. Notice the initial prompt ā€“ how I give it context. Later my interactions became more conversational (I was using the mobile app and simply talking to it as if I were a new lawyer). 

šŸšØšŸšØ Now I am not suggesting this is enough to replace mentorship programs. What I am suggesting (and proving in my own work) is that there is a path forward here to find a way to scale personalized, highly effective mentoring.

BTW: Parents are doing the same thing. Here is a great story of a parent who used ChatGPT to create a personalized math tutor for his daughter.

Spotlight

Total Recall: Microsoft Freaks User Out with AI-Powered Recall Feature for Windows 11

Microsoft's new Windows feature, Recall, can find anything youā€™ve seen or looked at on your PC, even if you donā€™t remember where or when.

This sounds cool but also really creepy and/or embarrassing. Listen, no judgment on what you look at on the internet, but do we want our computers to remember everything? Plus, letā€™s be clear here, BigTech does this not to just give us cool features but to monetize our privacy. This recall will allow more surface areas for directed and targeted ads.

So what is Recall:

  • Photographic Memory: Recall acts like a photographic memory for Copilot+ PC users, enabling easy retrieval of any past activity.

  • Marketing Insight: "Anything you've ever seen or done, you'll now more or less be able to find," says Yusuf Mehdi, Microsoft's consumer marketing chief.

How It Works:

  • Small Language Models: Runs on the PC itself for better performance, privacy, and environmental benefits compared to cloud-based models.

  • Search and Find: Users can search by typing details or use a slider to navigate the PCā€™s digital memory chronologically.

  • Direct Access: Retrieve images or text directly from Recall without needing to know the storage location.

  • Comprehensive Logging: Windows takes screenshots, decodes text in images, and logs meeting activities for easy rediscovery.

Privacy Controls: Users can delete specific snapshots, exclude certain apps or websites, and pause Recall temporarily.

Limitations:

  • Local Storage: Data is stored locally, meaning Recall only works on the machine itā€™s installed on and isnā€™t accessible from other devices.

  • New Machines Only: Recallā€™s memory starts fresh with a new Copilot+ PC and isnā€™t available on older models.

AI Model Notables

ā–ŗ Immigration law firm partners with legal tech company to provide law school clinics free access to cutting-edge AI technology.

ā–ŗ Microsoftā€™s Copilot embraces the power of OpenAI's new GPT-4o.

ā–ŗ Apple and OpenAI are expected to announce their partnership at next month's WWDC event.

ā–ŗ Google DeepMind launches Frontier Safety Framework to identify future harmful capabilities. The framework, set to be implemented by 2025, will assess AI systems at three distinct levels: the model's capabilities, human interactions with the technology and the broader systemic impacts. This approach aims to identify and mitigate risks associated with the deployment of advanced AI models.

ā–ŗ OpenAIā€™s May 16th trademark filing for GPT-4o frequently mentions agents, including ā€˜virtual, autonomous, semi-autonomous, and intelligentā€™ variations.

News You Can Use:

āž­ Scarlett Johansson says she is 'shocked, angered' over new ChatGPT voice

āž­ A new artificial intelligence task force will lead research and development on the use of generative AI in California courts

āž­ Colorado becomes first state with law regulating potential consumer harms of artificial intelligence

āž­ The highest-ranking judge of Australia's oldest court has flagged generative AI as one of the biggest challenges facing the justice system (deep fakes and interrogating expert AI systems)

āž­ Hollywood agency CAA aims to help stars manage their own AI likenesses

Was this newsletter useful? Help me to improve!

With your feedback, I can improve the letter. Click on a link to vote:

Login or Subscribe to participate in polls.

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.

DISCLAIMER: None of this is legal advice. This newsletter is strictly educational and is not legal advice or a solicitation to buy or sell any assets or to make any legal decisions. Please /be careful and do your own research.8