140 | 🐂💩 Bard's BS Detector

In today’s Brainyacts:

1. Dusting off an OG prompt

2. Bard’s BS detector

3. LinkedIn’s AI job finder bot and more AI model news

4. How do certain AI models do with legal-related tasks and other related news

👋 to new subscribers!

To read previous editions, click here.

Lead Memo

OG Prompt

With new readers coming in all the time combined with the realization that I know most people do not have (or make) the time to play with these tools, I am dusting off some oldy but goody prompts to reinvigorate you and inspire you.

This is for all of you. You all need a bio. Be it for your LinkedIn, intranet directory, or you are giving a talk at an event. You need a Bio. But most of us write boring bios. Worse, we reuse them without thinking. This is especially bad if you are giving a talk at an event. Pros edit their bios to fit the audience and the context. If you are not giving talks, you should review and update your bio frequently as you collect different experiences.

So, some so-called prompt experts would tell you to just ask ChatGPT to write you a bio based on some basic info you give it. That might work but it will be lackluster.

Instead, do this.

▶︎▶︎PROMPT

I need to update my bio. I am going to paste my resume and job description below. Based on this, I want you to interview me with questions that will help you write a concise, compelling, and credible bio. I need this for a talk I am giving to a law school’s faculty on the role of generative AI in the current and future legal education market. [paste resume and job description (if you have one) - oh and I always remove my name and address before pasting it in OpenAI]

Ok, see what I did here. I told it what my goal was - not just a bio but one that is concise, compelling, and credible. That will make a difference in how ChatGPT interacts and ultimately drafts a bio.

I also seeded it with context - my resume - but also my audience and topic. Now if you are doing this to update your LinkedIn bio you won’t need to do that. BUT YOU SHOULD tell it who you are trying to attract on LinkedIn. You have a Profile for a reason. Who is the single most important type of person you hope connects with you? Potential employer. A thought leader in your space. Potential job candidates. Tell ChatGPT this.

Here is an example of what it gave me: https://sharegpt.com/c/d4ixSRj

Now I have a number of questions to answer. There are many. And you know what? I don’t have to answer all of them. Remember, you are still the boss here, not AI.

I answered two of them:

It then gave me a long and detailed bio. I had to ask it to shorten it.

Here are the bios:

▼▼RESPONSE

Certainly! Here are two versions of your bio: one that is approximately 50 words and another that is around 100 words:

50-Word Version:

[Your Name] is a legal innovator, director of the Legal Innovation & Entrepreneurship Program at the University of Richmond School of Law, and co-founder of Bold Duck Studio. With a background in law firms and in-house roles, [Your Name] bridges the gap between legal education and real-world innovation, emphasizing the transformative potential of generative AI.

100-Word Version:

[Your Name] is a visionary legal professional who serves as the Director of the Legal Innovation & Entrepreneurship Program at the University of Richmond School of Law. With extensive experience in law firms and Fortune 500 in-house teams, [Your Name] brings a unique perspective to the evolving legal landscape. As the co-founder of Bold Duck Studio, [Your Name] collaborates with global law firms and corporate legal teams to drive innovation, change management, and strategic transformation. A sought-after educator and guest lecturer, [Your Name] advocates for incorporating generative AI awareness and practical skills into legal education, equipping students for the dynamic future of law.

Are these pure gold? It depends. Are they what I could have come up with on my own? Perhaps. But that is not the point. The point here is to see how I interacted with it. And yes, I do believe that it can help any of us rethink how to approach and communicate our bios.

Simple. Easy. Instructive.

Spotlight

Google Bard Detects Hallucinations

Bard was a bit of a joke when it first debuted. But over recent months it has gotten much better. A short bit ago they added a new feature.

Basically, you can now use the power of Google Search directly in Bard to stress-test Bard’s responses for accuracy and trustworthiness. In other words, Bard now has a bullish!t detector built in. And it works (mostly).

Check out my short 4min video.

AI Model Notables

► LinkedIn launches new AI chatbot to help you get a job

► Baidu (China) launches paid version of ChatGPT-like Ernie Bot

► Apple’s AI health coach

► HuggingFace’s Zephyr 7B Beta beats ChatGPT

► Microsoft releases big Windows 11 update with Copilot AI assistant included

News You Can Use:

➭ The Beatles’ final song is now streaming thanks to AI

➭ AI Safety Summit: China, US and EU agree to work together. See the complete guest list here

➭ ‘AI’ named most notable word of 2023 by Collins dictionary

➭ UN AI report to close gaps in government responses

➭ Next Senate AI forum will focus on elections

➭ White House unveils AI.gov in a historic move towards comprehensive AI oversight 

➭ Fascinating research: Rating Generative AI Responses to Legal Questions (results)

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.