No - I don’t think the technology is ripe for small firm attorneys (which I am) yet due to lack of privacy and availability.

Chat GPT4 is not confidential unless you buy or build your own enterprise version and train it on your data. Not practical for small firms. Windows Copilot is “generally available” to firms with 300 or more subscriptions at $30.00 per month. Copilot won't be available for other until at least May of 2024. There are some AI’s (LLMs) being employed at large firms (1,000+ staff) such as Harvey.AI but from what I have read it seems that there is a custom AI being built for each firm.

If you have used GPT 4 you will know that getting good output requires good, detailed, and structured input. The time it would take to draft the input for a good demand would for me exceed the time to write a demand. It would make more sense to create a template in smokeball to import this information from your smokeball matter data. What would reduce this gap for me is if there was an AI which could review the medical records or police reports etc., summarize them, and create the demand for you using that information. I’m thinking something like an extension of these:

Unlock data value with healthcare data solutions in Microsoft Fabric - Microsoft Industry Blogs
Azure AI Document Intelligence | Microsoft Azure
 

I would be interested to know what Smokeball sees the future holding for the software and AI integrations.

 

@msimowitz This would be a great topic to bring up during our AI Client Round Table on January 10. And thanks for your insightful response @JKibler!

@msimowitz, I have started a free trial of casetext’s cocounsel all access Link here. I got a same day demo when I called casetext. This can ingest documents, add OCR, and summarize them. It keeps your data confidential and does not train using your data. This might be a solution for your demand letter writing need. It also does excellent legal research based on one comparison of 2.5 hours of westlaw research by me compared to ten minutes by casetext.

Thomson Reuters purchased casetext in august for a huge sum. As happens after acquisitions, Westlaw is reportedly going to eliminate the top tier, $400/user per month plan and require the $250 per month plan plus Westlaw precision AI at their “variable rate” pricing beginning 1/1/2024 which may be as much as $700/user per month. I am told by the sales rep that if you subscribe prior to 1/1 then you can get the product for $400/user per month. I’ll update this comment or thread later after I complete the trial. 

I have added AI to my LexisNexis field, which eliminates the confidentiality and questionable sourcing issues with Chat GPT and Microsoft AI. The LexisNexis AI is strictly limited to case law, statutes, Practical Guidance and secondary content curated by LexisNexis, which are the same sources I use in researching an issue. The response also identifies the authorities used in the response. I am cautiously optimistic after one month of using the service.