I've picked up a few really exciting thoughts over the last 2-3 days that I wanted to share. The current software sale is part of a rotation that has been going on for a while now
The thesis: HALO instead of Asset Light 😇 Josh Brown (The Compound) put it aptly: for years, "asset light" (software, data, consulting) was the non plus ultra. High margins, no physical ballast. Today, that's exactly the risk. If your product consists only of information, an LLM can replicate it. The flight is now into HALO (Heavy Assets, Low Obsolescence) - companies that own "real" things that can't just be prompted away.
SAAS - Is the "per seat" model dead?
Nate (super exciting guy) posits the thesis: AI needs software. AI doesn't just replace $CRM (-0,08 %) or $SAP (-0,29 %) & CO. That's nonsense. But that's what SaaS is struggling with. AI replaces people >>> The "per seat" business model, the non-plus-ultra of the SaaS world, is faltering.
Interesting sectors according to Dan Ives (Sell Side - has funds himself, bias!!!)
Enterprise AI: Palantir ($PLTR (-0,84 %))
"The first address" for AI applications in government and companies. He sees the path to a trillion-dollar valuation here.
Cybersecurity: CrowdStrike $CRWD (+0,46 %) & Palo Alto $PANW (-0,02 %)
AI makes attacks more dangerous, so every company needs better defenses.
Data layer: Snowflake $SNOW (+0,46 %) & MongoDB $MDB (+0,23 %)
Before an AI can work, the data must be cleanly structured. These companies are the "garbage collection and sorting system" for AI data.
Platform giants: Microsoft $MSFT (+0,23 %) & Salesforce $CRM (-0,08 %) or $SAP (-0,29 %)
He sees Salesforce as massively undervalued (P/E ratio of approx. 15), as the market is ignoring the potential of "Agentforce" (AI agents).
AI models are worthless without context. The big SaaS players are sitting on the companies' historical data. Whoever has the data controls the AI agents that access it. That's the reason why, according to Ives, they are not simply "prompted away".
Sources:
Okay Computer Podcast (Dan Ives / Dan Nathan)
The 200-line prompt that killed $285B (NoteGPT/Anthropic Analysis)
What Are Your Thoughts / Animal Spirits (Josh Brown & Michael Batnick)
