The speaker argues the market is in an AI-led everything rally, with the most attractive opportunities still in compute, data centers, and select AI infrastructure names like Nvidia, CoreWeave, Nebius, and Micron. He is notably more cautious on software and says many SaaS businesses may face margin pressure or contract renegotiation as models improve and agents can replicate more tasks.
Watch on YouTube ›Get the market thesis, key claims, assets, contradictions, and follow-up questions from any financial video — then unlock a version personalized to your portfolio, watchlist, and favorite speakers.
This was a fast-moving midday market wrap centered on the speaker’s core thesis that the current rally is being driven by AI infrastructure and is still early enough to favor select compute-linked names, even after large gains. He repeatedly framed the environment as an “all rally” for tech, citing sharp moves in Nvidia, ARM, CoreWeave, Nebius, Oracle, Cadence, and Micron, and argued that the market is rewarding companies with direct exposure to GPU demand, power, data-center capacity, and engineering excellence. The main bullish case was strongest for CoreWeave. He argued that CoreWeave offers better risk/reward than Nebius at current prices because it has faster growth, a lower market cap, and more differentiated infrastructure/software. …
The near-term setup is still momentum-friendly for AI infrastructure, but the trade is getting crowded fast. I would treat new entries as tactical, not casual, because the next move may be more about rotation and pullbacks than fresh discovery.
Over the next few weeks or months, the base case is continued capital spending into GPUs, data centers, and adjacent enablers if frontier-model demand stays hot. The key question is whether CoreWeave and similar names can keep converting that demand into margin and contract durability, or whether the market starts to anticipate an efficiency inflection.
Longer term, the speaker is arguing for a regime where AI compute becomes a core industrial input and software pricing power becomes less reliable unless tied to network effects or embedded workflows. If that regime holds, the durable winners are likely to be the technical infrastructure providers with the best execution and access to frontier chip supply.
The market is in an AI-led tech rally, with many infrastructure and software names surging together.
He opens by describing an all-rally environment and lists multiple winners across AI and tech.
CoreWeave offers better risk/reward than Nebius at current prices.
He explicitly says CoreWeave has more upside now on a risk/reward basis than Nebius.
CoreWeave's growth is being under-modeled by Wall Street and could reach about $5 billion revenue by March 2027.
He cites Wall Street estimates and says the company is growing faster than modeled.
What market cap do you see Nebius and CoreWeave getting by 2030?
Tanner says he hasn't given that specific thought, then pivots to discussing Leopold Aschenbrenner's thesis about superintelligence driving massive data center scaling from 100MW to 1GW to 10GW to 100GW, noting that CoreWeave and Nebius could be worth hundreds of billions if model improvements continue but face risk if a data wall hits.
How do you power 100 gigawatts of data center capacity?
The speaker notes it would be about 20% of total US energy consumption, then returns to discussing valuations of CoreWeave at 400 billion by 2030 and comparing Nebius' growth story.
What are your thoughts on Galaxy given they got a lot of power approved?
The speaker says they're not sure and that they're buying the companies with the best data centers, not the most power, effectively declining to give a substantive answer.
Unlock the full claims, asset map, scores, related transcripts, follow-up questions, and AI chat — shaped around your portfolio, watchlist, favorite speakers, and risks.
