Is GenAI a bubble ready to pop?

It is entirely possible that GenAI is both a useful technology and a bubble waiting to pop

This is a thought that's been niggling away at me for the past 12-18 months, and it was the LegalTechTalk conference last week that crystallised it for me

The conference itself was fantastic, and all the more commendable for it being in only its second year

But there are a few warning lights flashing an ever darker shade of amber:

1. The volume of legal tech start ups is truly mind boggling. I always recommend Legaltech Hub as an invaluable resource to help here, but I do not feasibly see how the typical law firm is expected to wade through the number of vendors to separate the wheat from the chaff

   
   

2. Good chat with Edward Wilson around Total Addressable Market (TAM). Given the number of vendors pitching for funding, even if we assume they are conservative in assessing TAM and their market share, I think it's all but guaranteed that the aggregate TAM and market share across all of those pitch decks is somewhere just north of ridiculous. This sets the stage for an almighty crash

   
   

3. "AI" is an increasingly vague term. A lot of the heavy lifting in good products is being done via less glamourous RPA, workflow and document automation - I don't recall referring to these as AI in previous hype cycles(?)

   
   

   And sure, these tools may have a GenAI component in the UI - but in the rush to appear hip and trendy that GenAI capability appears oversold. My concern here is that when we truly hit the trough of disillusionment with GenAI it's going to have a severe chilling impact on these other technologies by association. Note to self here - be more specific in identifying and describing the tech being used

   
   

4. There was an F1 car at a Legal Tech conference. No, I don't know why either 🙂. To a conversation with Daniel Pollick, if that's not an indicator that we might - just might - be in bubble territory, I don't what signs we're looking out for...

   
   

5. Stepping back from GenAI, statistically most VC backed companies fail (75% per HBS). Whilst the founders and VC backers know what they're letting themselves in for, this is a relatively new dynamic for law firms and buyers of this type of tech more broadly. What's the plan for when your preferred legal tech product goes pop?

We can put to one side evangelism (or not) for the underlying technology here. Even if firms are bullish on the underlying tech, serious consideration must given to (a) the longer term commercial viability of the vendor, (b) monitoring of key vendors and (c) contingency planning if a particular vendor fails

[Image generated with ChatGPT - would it even be a post about AI without one? 🙂]

Iamge of a light blue, F1 car balloon on string