Take-Two’s thesis shifts from GTA monetization and open-world cash-cow upside to an AI-driven valuation problem: if game creation becomes cheap and fast, legacy publishers could face multiple compression even if they keep shipping strong titles for 5–7 more years. GTA’s expected hit status is treated as already priced in; the real question is scale, with 30 million units cited as the baseline and 130 million as the level that would matter to the stock. The position was cut hard after that narrative change, down to roughly 20%–25% of the prior size, while keeping a long stake. The argument frames the risk as a 2035 problem: if tens of millions can generate games on the fly for near-free, revenue durability for TTWO and peers weakens sharply.
AI acceleration over the past couple of months has increased investor fear that game development could be democratized sooner than later, creating inevitability risk for TTWO.
GTA’s being a hit was already expected, so “hit” alone is a nonissue; only the scale of the hit matters to TTWO’s thesis.
A 30 million-unit GTA projection is cited as the market’s baseline; 130 million units would be the level that materially changes the stock story.
The prior dominant TTWO narrative was GTA monetizing the open world over the next few years and becoming a cash cow, possibly 2x bigger than the market expected.
The narrative has shifted from GTA monetization to concern about what other games TTWO and other gaming companies can make.
The core valuation question is whether there is still value in legacy big publishers in 2035 if anyone can create a game on the fly.
Even if legacy publishers can make amazing games for 5–7 more years, AI-driven content creation arriving in 5, 6, or 7 years could compress their valuation multiple.
The comparison is to SaaS multiple compression: CRM is not going away, but AI-enabled substitutes could drag on growth over the next 20 years.
Incumbents with a 15- or 20-year history of bug fixes still matter, but AI-era startups can use cheaper competitors coded for next to nothing.
The future gaming model is framed as tens of millions of people creating games on the fly, with very small communities playing them essentially for free.
If gaming shifts toward AI content creation, long-run revenue generation for TTWO becomes doubtful and valuation risk rises.
People are back-engineering AI tools, and similar workflows are suggested for GTA 6 shortly after release, possibly in a few days.
That few-days claim is pushed back against because the game engine does much of the work, with the timing compared loosely to Adobe-suite work.
The TTWO position was reduced meaningfully after the narrative change, but not closed; it remains a long position.
The reduction is described as roughly 20% of the prior position, maybe 25%, implying about 75% was sold.
Selling most of the TTWO position after the thesis shift is presented as avoiding an enormous amount of loss as the stock moved.
A key unresolved question is whether GTA 6’s launch will prove AI can replicate major game workflows quickly enough to matter to TTWO’s multiple.