065 | Giving Hugs

Brainyacts #65

My so-called “Big Discovery” yesterday

Yesterday, while working in GPT-4, right before 10amET, I thought I discovered something and was like . . . .

I quickly made a video to share with all of you. I even titled yesterday’s newsletter referencing this discovery. 😧😬

At around 10pm that night, as I was lying in bed about to fall asleep, I began second-guessing my great discovery . . . something was gnawing at me . . .

When I woke at at 5am and saw my inbox, I was immediately in a state of this 👇!

Um, yeah. So, I have the best readers one could have. After sharing what I thought was an amazing GPT-4 discovery yesterday . . . lo and behold I was just being a bonehead.

So no, GPT-4 did not remember me.

In trying to bring you the best of the latest and greatest every day, in this case, I got a bit too excited and saw a mirage (or didn’t see something staring me in the face).

Many of you reached out to let me know but also encouraged me and thanked me for all the accurate and helpful insight I am sharing. Thank you!

Well, I guess after 64 editions, having one embarrassing oversight is allowable. I am giving myself some grace and moving on. ⏩ 🚀

Let’s keep climbin’.

A special welcome 👋 to my NEW SUBSCRIBERS! 

To read previous posts, go here.

In this edition we will

  1. introduce a new chat tool

  2. meet Emma Haywood

  3. shake our heads at the Dept. of No

  4. give the Dept. of No a plan forward

  5. make resume and cover letter writing less awful

  6. catch up on EU AI policy news from today

  7. shrug at the unimaginative Wall Street Journal

  8. get a 15-second recap of the Google event yesterday (meme-style)

New Chat Tool HuggingChat

HuggingChat has been released. This is a new chat-driven tool, similar to ChatGPT, that you can try for free. It is still a bit quirky, but worth some time to experiment. See one of my test runs here.

This is a free and open source chat assistant that is designed to give you similar functionality to paid services like ChatGPT. With HuggingChat, you have more freedom to choose the models, features and customizations you want, although the stability and uptime guarantee may be lower.

In this v0.1 of HuggingChat, users are not authenticated in any way, i.e. this app doesn't have access to your user account even if you're logged in to huggingface.co. The app is only using an anonymous session cookie.

❗️ Warning ❗️ this means if you switch browsers or clear cookies, you will currently lose your conversations.

Things I find interesting:

  • No login required

  • The results seem faster than ChatGPT

  • You can share direct links to conversations like I did above

  • The quality is unpredictable. Sometimes it seems to perform just as good as GPT-3.5 (Free ChatGPT) other times it goes a bit bananas 🍌 like in what I shared above.

  • You can switch between models. There's one for asking technical questions. But switching models is not for casual users.

Shout-outs!

Meet Emma Haywood everyone. Emma has driven 10+ subscribers to this newsletter so she gets an exclusive shout-out.

Below is Emma in her own words. She is up to some truly interesting things and I LOVE that she is an entrepreneur - venturing out to build her own business.

Go Emma! She is based in London.

Hi Josh,

This is great - thank you! Yesterday I spoke at an event on ChatGPT and generative AI for a group of in-house lawyers. I shared a link to The BrainyActs as follow-up, and I see it was popular!

Some more information about me:

I’m a London-based commercial/IP/tech lawyer. I’ve just launched my own independent legal consultancy, Bloomworks Legal (https://bloomworkslegal.com/), offering outsourced in-house counsel services to purpose-led businesses and innovative cultural institutions.

Beyond the law, I have a keen interest in the 4-day work week. I write a monthly newsletter called The 4 Day Lawyer (https://4daylawyer.substack.com/) and take on consulting and mentoring projects to promote alternative working patterns within the legal profession. I’ve recently been thinking about whether generative AI will bring us any closer to a universal 4-day week with no reduction in pay (which is the ultimate aim of the 4-day week trials you may have read about in the news). I’m currently writing the next edition of my newsletter on this subject - hopefully out on Monday.

Also, [here is the] most recent edition of my newsletter on generative AI and the 4 day working week: https://4daylawyer.substack.com/p/4dw-x-chatgpt-and-an-announcement

Thanks!

Emma

The Department of SLOW and NO: 

It’s The Business, Stupid.

In 1992, Bill Clinton's campaign strategist, James Carville, famously coined the phrase, "It's the economy, stupid," a blunt yet effective mantra that went on to define the presidential race.

Today, as we wade through the tides of the modern legal profession, a similar awakening seems to be upon us (well one I am hammering on). If Carville were to materialize in our boardrooms, he might be seen scribbling a slightly tweaked version of his famous line on our whiteboards: "It's the business, stupid."

No disrespect to the esteemed lawyers among us, but let's face it: there's been a seismic shift in the legal landscape. Our legal departments are no longer just the gatekeepers of compliance or the bearers of bad news (you know, the ones who show up to say "no" just when the party gets going). In our fast-paced business world, where legal intricacies entwine with commercial operations, the legal department is expected to be a proactive partner, a dynamic player in the game of business strategy.

But apparently the general’s counsel’s desire to be at the head table of corporate vision and strategy is one based on vanity rather than conviction. At least that is how I read most times.

Legal’s Customer Service Scores Keep Getting Worse

In the rapidly transforming business landscape, the legal department often seems like a vestige of the past, struggling to keep pace with more nimble counterparts. However, the concept of generative artificial intelligence (AI) holds the potential to revolutionize the legal field, providing a means to address many of the issues raised by internal business customers. Yet, this potential remains largely untapped. Why is this so, and how can we bridge this gap?

The recently released 2023 Enterprise Legal Reputation Report by Onit Inc. provides compelling insights into the dilemma. The report highlights the increasing disconnect between legal departments and other business units, with a lack of efficient communication and response time being the primary culprits.

Just look at the decline satisfaction over just one year! This is not a good look for in-house legal teams.

While legal departments are still seen as the business's trustworthy protectors, their reputation for bureaucratic inefficiency is causing frustration and non-compliance among non-legal staff.

This paints a grim picture, but it also offers an opportunity.

Enter generative. You all know this as readers and AI enthusiasts. However, the adoption of this technology in the legal field has been slow, largely due to budget constraints, lack of technological solutions, and frankly a complete disregard for running legal as a business rather than a self-righteous guild.

Despite these challenges, generative AI's potential for transforming the legal department into a more efficient, responsive, and collaborative entity is too vast to ignore.

Here are some ways that a corporate legal team could leverage ChatGPT/GPT-4 to address many of the issues raised by their internal business customers. I am including ideas that individuals could take action on to prototype and derisk the idea a bit to help get things moving.

  1. Improving Communication: Generative AI can automate routine communications, providing quick and concise responses to simple inquiries, thereby improving the department's responsiveness and freeing up time for more complex tasks.


    🙋‍♀️An individual could use ChatGPT/GPT-4 to prototype an AI-driven communication workflow. For instance, they could develop a simple ChatGPT/GPT-4 prompt template that they use to respond to routine legal inquiries from other departments. Run it manually to test it with a set of frequently asked questions to ensure accuracy and appropriateness.

  2. Enhancing Efficiency: AI can analyze and process vast quantities of data at a pace that humans can't match, significantly reducing the time taken to perform tasks like contract review or legal research.

    📑A proof of concept could involve using ChatGPT/GPT-4 to summarize legal documents (SOWs, MSAs, etc.) or extract key points from large volumes of text. The individual should ensure the AI system's output is cross-verified by human reviewers in the early stages to manage risk.

  3. Facilitating Contract Lifecycle Management (CLM): AI can streamline the contract management process, making it more efficient, accurate, and speedy. This would lead to faster deal closures and revenue generation, addressing one of the main pain points identified in the report.

    🔂Using ChatGPT/GPT-4, an individual could prototype a prompt that can review and flag issues in contracts. The prototype could be tested on past contracts and the results compared with human reviews to assess accuracy. Any use of the tool in live contract reviews should be supervised until confidence in its accuracy is established.

  4. Cultivating Collaboration: Generative AI can facilitate better collaboration between departments by providing a platform for more streamlined communication and efficient processes, effectively breaking down silos.

    🫸🫷🏽One of the chief complaints by other company functions is that Legal is not user-friendly and speaks its own language - even in emails. An easy prototype using ChatGPT/GPT-4 would be to create a prompt that summarizes legal-heavy text and translates it into plain-language. Better yet, ChatGPT/GPT-4 could also convert long texts to bulleted info for easier consumption.

  5. Assisting in Compliance: AI can also help in ensuring compliance by automatically scanning documents for potential regulatory issues and flagging them for review, reducing the risk of bypassing legal procedures.

    The individual could use ChatGPT/GPT-4 to create a prompt that scans documents for compliance issues. The proof of concept should be tested on a controlled set of documents, and the results cross-checked by a legal expert.

However, to fully realize these benefits, corporate legal teams must overcome their reluctance towards adopting new technology, experimenting without a fully rolled out technology plan, using pilots or prototypes, and overcomplicating all of this by making efforts fully-baked executive-backed committee-approved projects - those take forever to start and seem to fad away after a few weeks.

💡🎁 ChatGPT/GPT-4 and related tools are a gift to in-house teams. Quit debating whether it is safe to use or not. There are innumerable ways of testing it out without any issues.

And BTW, talk to your business colleagues about it as I am sure they are using it and testing it too - you just don’t know because they are afraid you will shut it down and tell them no!

USE CASE Getting a Job

Writing cover letters and reworking your resume is never fun. Neither is prepping for interviews - trying to think of all the questions (usually generic) that somehow you have to come up with crafty answers that are somewhat sincere.

Cue the “what’s your greatest weakness?” question.

My greatest weakness is that I will drop dead at my desk working my butt off for you.

Hired!

All kidding aside. Here are some tools that will write, review, strengthen, and prepare your materials and talking points:

  1. MajorGen: A tool to create resumes and cover letters from profiles.

    Link: https://www.majorgen.com/

  2. Cover Letter AI: AI tool that helps write professional cover letters

    Link: https://coverletter-ai.com/

  3. Kickresume: Create a resume quickly with the help of artificial intelligence

    Link: https://www.kickresume.com/en/

  4. Resumecheck: Optimize resumes for specific job roles

    Link: https://resumecheck.net/

  5. Applai: A platform for job preparation and cv optimization.

    Link: https://www.applai.me/

If you want to use ChatGPT instead, here are some prompts to use, refine, and adjust:

▶︎▶︎PROMPTS

[ ] = where you might want to add specifics

Please review my resume below and provide detailed/constructive feedback on its overall format/structure/design, highlighting areas that could be improved/optimized for maximum impact? This is being submitted to [name how and to who – such as HR person, the hiring manager, or via online resume submission]

How can I write a standout/impressive resume/CV that showcases my unique [skills/experience] to [potential clients/employers] in my [industry/field], and effectively communicate my [value proposition/competitive advantage]?

In what ways can I tailor my resume/CV, pasted below, to emphasize my strongest [skills/achievements] for my target [industry/field], while de-emphasizing areas that are less relevant or may not be as valuable to [potential clients/employers]?

What specific [keywords/phrases/industry-specific jargon] should I include in my resume/CV below to help it get noticed by [potential clients/employers], and how can I optimize it for [the specific job or project description] you are applying to?

Please review my resume and tell me how can I effectively highlight my [achievements/accomplishments], using specific metrics/results/outcomes to demonstrate my impact/value to past employers?

News you can Use: 

EU Takes More Action on Regulating AI

In a move towards tighter regulatory control over artificial intelligence (AI), the European Union's internal market and justice committees have agreed that companies creating fundamental AI models should be required to:

  • produce risk assessments,

  • summarize copyrighted materials used in training the models, and

  • ensure users are aware when interacting with AI.

The legislation, termed the Artificial Intelligence Act, aims to provide global influence and begin worldwide discussions on AI regulation.

Key Statements, Developments, and Takeaways:

  • The committees also voted to ban the real-time use of AI for identifying people in public.

  • The voting results on the EU’s Artificial Intelligence Act showed the majority of lawmakers (84) in favor of the additional controls, seven against, and 12 abstentions.

  • This vote is not final; the entire European Parliament plenary will vote on the full AI Act proposal next month, and then the parliament will negotiate a final deal with the commission and the EU’s 27 member states.

  • The authors of the AI Act aim to influence other governments worldwide to implement similar regulations on AI technology.

Upcoming Key Dates or Events:

  • The entire European Parliament plenary will vote on the full AI Act proposal next month.

  • Sam Altman, OpenAI CEO, is set to testify in front of the US Senate next week, addressing the subcommittee responsible for regulating developments in AI.

News you can Lose: 

Wall Street Journal Uses Clickbait Headline

I find it so tiring that major media just cannot help themselves when it comes to covering the legal market. It is almost always something akin to “the death of the billable hour.” Sure I get it - few like the billable hour but the power of generative AI is much greater than relegating it to just the economics. How about client satisfaction, lawyer work-life improvement, unlocking business and operational excellence, and so on. They just miss so many better angles.

  • Big law firms are testing artificial intelligence (AI) tools, such as GPT-4, to handle tasks typically assigned to entry-level lawyers and streamline complex work.

  • These AI tools can perform writing and research tasks, such as drafting documents, analyzing contracts, and conducting legal research at a much faster pace.

  • The incorporation of AI in the legal field could potentially reshape white-collar occupations and the practice of law itself, according to experts.

  • AI startup, Latch, uses GPT-4 to simplify the contract review and redlining process. It launched in early April and has a wait list of over 80 companies, including law firms and in-house counsel.

  • Global firm Allen & Overy uses GPT-4-based tool, Harvey, for tasks such as legal research, drafting documents, and contract analysis. It has resulted in lawyers spending less time on difficult tasks and more time on core tasks.

  • Companies like Microsoft and Ford Motor have beta-tested a product called CoCounsel, which uses GPT-4 technology for preparing depositions, contract analysis, and legal research. Notable law firms, including DLA Piper and Kirkland & Ellis, were also part of the beta group.

  • Despite the advantages, some clients express concerns about the security of cloud-based products and the transparency of AI software.

  • Legal research and analytics provider LexisNexis has also created a platform utilizing GPT technology.

  • The adoption of AI could potentially change the revenue model of the legal profession, making the traditional billable hour less prevalent as AI reduces the need for some associate work.

In the Memetime: 

If you missed the Google event yesterday, here is a 15 second recap to catch you up:

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.

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.