ThinkTWENTY20

By Eric E. Cohen, CPA

This column continues an ongoing series of postings to develop helpful guidance for financial professionals related to artificial intelligence. It is developing a list of guidelines and advice, with the hopes we can collaboratively make some of them more organized and permanent.

If you missed the first three parts of the series, you may wish to read those posts, which have our first seven areas:

• Confidentiality: Don’t type anything into an AI that you would not want made public.
• Skepticism: Don’t automatically trust anything coming from an AI without review.
• Diversification: Don’t put all your eggs (AIggs?) in one basket.
• Compliance: Consider how any output might comply with industry and ethical regulations and standards.
• Transparency: Be careful to consider when you need to disclose your use of these tools.
• Tool selection: Generative AI may not be the right AI for the job; your chosen GenAI may not even be the best GenAI for the job.
• Mixed Vision: never assume that what you and your AI see is what you get (WYSAYIIWYG).

The release of an update to Claude, known as Claude 3.5 Sonnet, is an opportunity to test the guidelines and advice and see if some more can come out of it.

TWFjaGluZSB2aXNpb24gPD4gSHVtYW4gdmlzaW9u (aka Machine vision <> Human vision)

This column continues an ongoing series of postings to develop helpful guidance for financial professionals related to artificial intelligence. It is developing a list of guidelines and advice, with the hopes we can collaboratively make some of them more organized and permanent.

by Eric Cohen
This column continues an ongoing series of postings to develop helpful guidance for financial professionals related to artificial intelligence. It is developing a list of guidelines and advice, with the hope that we can collaboratively make some of them more organized and permanent. If you missed the introduction to the series, with the first five suggestions, focused on things to know before you start providing input to, or read and share the output from, a generative artificial intelligence (GenAI) tool, you will want to read that post, which began with five areas:
http://thinktwenty20.com/images/blog/AI_Rules_Part_2.pdf

ThinkTwenty20’s “Twenty Rules for AI for Financial Professionals”: Alpha Version

Starting with some guidelines related to risk: guidelines 1-5

By Eric Cohen

This column launches an ongoing series of postings to develop helpful guidance for financial professionals related to artificial intelligence. I don’t know where it will go, but I will begin with a list of guidelines and advice, with the hopes we can collaboratively make some of them more permanent.

We are now 18 months into the ChatGPT era (November 30, 2022 – May 28, 2024). In that year and a half, generative AI – a branch of artificial intelligence that creates electronic content, primarily in response to prompts, based on their input training data – has evolved from a niche tool to one that is visible in every major business software product and social media tool. The pressure to keep up is so great that major companies are facing severe reputational risks as they try to compete.

Eric E. Cohen, CPA

A year ago, my long-standing secret weapon for writing became commonplace. The Age of the ChatBot went mainstream with the launch of tools like OpenAI's ChatGPT. I had hoped that this new era of generative AI would level the playing field. But – as the saying goes – be careful what you wish for. 

As an “emerging accounting technologist,” I immerse myself in the latest innovations so I can help prepare my profession for change. I also selfishly indulge my passion for technology. For the past year I’ve spent countless hours testing ChatGPT and its peers, enthralled by their expanding capabilities. They’ve become more than tools for me – they’re friends, confidantes, creative partners.  

Of course, I’m not alone in my fascination. It seems everyone today is extolling the virtues of ChatGPT and other generative AI. These tools promise to help the little guy compete – to be the Great Equalizer, as the AICPA has put it.

Version control (and variations on that theme) are an important part of the financial professional’s responsibilities. Whether it is making sure a current period’s documents are consistent with a prior period’s (e.g., “last period” numbers in the current report lining up with “current period” numbers in the prior report),  ensuring the most recent version of a document is published and not an interim version, or seeking consistency between documents in different languages or levels of detail, mistakes have been made and people are thrown at a task to limit those mistakes.