$NBIS bulls don’t need Gartner to be wrong. They need Nebius to survive the shakeout, and grow on terms that reward shareholders Gartner’s Oct. 14 forecast puts two things together: 1. Over 80% of neoclouds could cease operating independently by 2030. 2. Spending on AI cloud infrastructure rises from $66B in 2027 to $174B in 2030. That includes consolidation. It isn’t a forecast of 80% bankruptcies. And Gartner warns that even this spending growth may leave most AI compute providers unable to recover their investment before the current generation of chips is replaced. A bigger market doesn’t settle who earns a return. Buzzberg’s tracked speakers are making that distinction: * @ren_stocks agrees with the shakeout. His $NBIS case rests on customer prepayments, pricing and access to funding. He still flags financing the next expansion. * @Sandeman52 doubts the forecast, but expects $NBIS to thrive even if it happens. On Oct. 14, he said it was the only neocloud he owned. * @moninvestor expects smaller providers to be acquired. His Oct. 15 response singled out $NBIS and $IREN. Ren and Sandeman disagree on the industry forecast. They converge on Nebius. Separately, during the Oct. 13–14 selloff: - $NBIS: @babyfolio reported adding at $198.50 on Oct. 13; @EndicottInvests reported buying more on Oct. 14 - $IREN: @jiahanjimliu reported averaging in on Oct. 13; @BlackPantherCap disclosed adding that day. Those buyers sit within a broader pattern. Across Oct. 8–14, 2026, Buzzberg tracked bullish views from: $NBIS: 35 distinct profiles $IREN: 16 $CRWV: 9 $NBIS led this three-stock comparison. Each profile counts once per stock; these aren’t purchase counts. The financing risk is getting harder to ignore. Reuters reported on Oct. 15 that CoreWeave’s April-issued bonds due in 2031 were yielding near 16% That’s a secondary-market yield, not a reset of the fixed interest rate $CRWV pays on those bonds. For the next buildout, the terms still matter. Customer contracts need chips, power and buildings. Interest costs reduce profits and new shares can dilute existing shareholders. TL;DR: You can be right about AI demand and still underestimate the cost of delivering it. Of $NBIS, $IREN and $CRWV, which has the strongest case for funding growth without giving away too much of the upside? Follow @buzzberg_ai for the dated calls behind the headlines - who’s adding, who’s getting cautious, and the evidence changing their minds.

Big Tech could profit from AI even if someone else builds the best model. That’s @AlmaCap114204's case: model leads can disappear. Customer relationships/distribution is harder to replace. On Oct. 14, he said he “would be long” four names: • $META: Facebook, Instagram and WhatsApp bring the audience; ads can fund AI usage. • $GOOG: Gmail, Drive and Docs put AI inside existing workflows. • $MSFT: Office brings distribution; Azure serves enterprise demand. • $AMZN: Bedrock and inference infrastructure can benefit across competing models. His memory calls give us a reason to follow the argument. On April 30, he argued demand had shifted while memory supply couldn’t catch up. From that day’s close through Oct. 14, $MU gained ~76.8% and $SNDK ~28.0% He still sees a moat in high-bandwidth memory. He’s also bullish on AI while passing on standalone frontier labs at their valuations. The connection: he’s looking for advantages competitors struggle to copy. Buzzberg tracks the calls behind that judgment. Follow what he backs, what he passes on, and why 👇

Your next trade idea could come from someone you don’t follow. Watch 45 seconds of Buzzberg Feed: trade ideas only. Filter by sector or market cap. Check the reasoning, source and price at time of thesis versus now. Then try your sector! Which view hasn’t reached your timeline? 👀

Three January ideas from @PhotonCap, each followed by a stock-price gain of more than 150% The research spanned AI PCs, AI networks and satellite laser links: $SNDK +320.7% | Jan 6 Photon highlighted the new Optimus SSD brand for AI PCs and workstations, and said they were continuing to hold. $LITE +150.5% | Jan 6 Photon highlighted lasers linking satellites, with Lumentum a potential beneficiary. $MRVL +220.7% | Jan 10 Photon saw the Xconn and Celestial AI deals strengthening Marvell’s copper and optical connections for AI networks. Selected stock-price gains from the first regular close after each post through Oct 12, 2026. Excludes dividends and fees; these aren’t total portfolio returns. Those were the January ideas. What has Photon Capital been researching since? 👇

Even a long-term $IREN bull is cautious about the @AnthropicAI headline. * @moninvestor is bullish through 2030. On Oct 12, he said he expects the reported deal to be small, with less impact on the stock than some expect. Bigger contracts could follow if IREN delivers. * @michaelsikand expects any Anthropic-driven pop to fade as investors weigh the buildout costs and dilution. The growth story matters. So does the bill. Before buying the headline, what would you need to see in the contract and how it’s funded? ↓

Anthropic @ reported $2T comes down to two questions: how much revenue and how much competition? Buzzberg’s dated investor views reveal the tension: 1. What will 2026 revenue actually be? CNBC reported a ~$65B annualised revenue run-rate at July-end. $2T is roughly ~31× that run-rate → @altcap | Bullish | Sees a $100B+ revenue trajectory, with 3–5× growth possible in 2027. Says @AltimeterCap would buy the IPO at scale → @PatrickEBoyle | Avoid | Argues the reported $2T valuation makes buyers pay heavily for growth that still has to happen 2. Can Anthropic stay ahead of OpenAI and open source? → @BillAckman | Watch | Maybe the best business story he’s seen. His hesitation: can frontier labs defend their lead against open models? → @chamath | Cautious | Praises execution, but expects open-source pressure to help force a lower IPO price → @goodalexander | Avoid | Prefers OpenAI’s product. Flags enterprise customers cutting inference spending and the risk of churn Even enthusiasm comes with a price condition. See the reasoning behind every move and the original sources linked 👇

Broadcom plans to lend Anthropic up to $42B to use its chips $AVGO would be supplying the chips and financing the customer. Two very different reads: • @babyfolio is bullish - the financing fears are overdone • @edzitron is avoiding - he sees too few big customers and more financing exposure/risk Buzzberg, past 7 days to Oct. 2nd: 8 bullish voices & 4 bearish voices These are views, not a list of "open" positions Before you pick a side, see what the bulls are counting on, and what the bears think could break Cases 👇

.@OpenAI's DevDay could benefit its rival, $META Our MCP surfaced the case - plus the arguments for $NVDA, $MU, $SNDK and $LITE Here’s why these five names came up, and what to check in your portfolio 🧵

AI needs more than GPUs. Who’s supplying the rest? Lasers. Memory. Storage Go deeper. Get the short version. Or just the one takeaway worth remembering Buzzberg MCP has you covered, right inside Codex Try it on a ticker you’ve bookmarked Setup 👇

Before you follow the money, follow the attention $NVDA tops Buzzberg’s most-talked-about list Big name tag team with a bigger crowd: 63% above its 30-day average Top mentions combine average attention and attention change See what other ticker stands out! Link in the first reply 👇

Anthropic signed a $11.5B deal with Akamai But the more interesting part is who receives @Akamai's money? Akamai estimates roughly ~$5.5B in capital spending to support the deliver the @AnthropicAI deal FinX's @aleabitoreddit take? He is looking at who gets paid i.e, hardware suppliers to $AKAM His potential beneficiaries, surfaced in Buzzberg’s Sept 24 Daily Alpha: • CPUs: $ARM, $INTC, $AMD • Memory: $MU, Samsung, $SKHY • Component procurement: $JBL His thesis: Follow the money and hardware needed to turn Anthropic's commitment into working compute To know stocks that deserve to be on your watchlist, but haven't appeared on X - check out Buzzberg's Daily Alpha: buzzberg.ai/daily-alpha?utm_…