Posts  / CAI  / #POST-248087
REDDIT

Is building AI overvalued and leveraging it undervalued?

C
Aug 21, 2026 · 17:22

For awhile most of the hype has on the buildout, the hyperscalers and obvious beneficiaries like security. **Does every company associated with AI already have that future priced into their stock?** Is there any convex upside worth investing in still? Won't leveraging AI be a larger market long term?

It’s also obvious that AI isn’t magic. It’s **great at some things** and **too unreliable** at others. But it is really good at processing and gaining insights from large amounts of information. 

So I ask,

1. Where does **applying AI create unusually large financial value**?  **Who owns own huge amounts of proprietary data**, valuable workflows etc that become substantially more valuable especially **as applied AI gets cheeper**, better and better managed/mitigated?
2. Which have the most **unrealized value** in forward earnings?
3. Which haven’t seen that value **already priced in**…or at least have further upside from where they have.

**Here are some I identified (with help from AI!). Several have already rerated so clearly this is not a new thesis.** 

**CAI** Proprietary cancer dataset, AI for treatment and drug development. I got in at $15.79 but it continues to rise.

**CCC:** Insurance/repair datasest and workflows

**Waystar (WAY):** Healthcare-payment dataset. AI helps prevents denials and recovers money

**Varonis (VRNS):** Agentic AI enterprise data permissions, though what stops a Microsoft from replicating

**Schrodinger (SDGR):** Physics AI for drug discovery. I got in at $15.35 but am unsure how much these PhDs can commercialize. 

**Sophia (SOPH):** Genomic/clinical data , AI for medical decisions. I got in at $5.91

**What do you think of these or others that might meet all three criteria?**

I also looked at but passed on, mostly because the future value is already priced in:

APP (had, then sold), TEM, HTFL, INTA

I did buy some NOW into earnings but unsure I'll hold it.