By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
Panelists generally agree that Alphabet's significant AI investments, particularly in Cloud, are risky due to uncertain ROI and potential regulatory headwinds, despite impressive growth and backlog. They express concern about negative FCF and the potential for AI to become commoditized, leading to a bearish consensus.
Risk: Uncertain ROI on massive AI investments and potential regulatory risks, such as forced Search unbundling.
Opportunity: Potential acceleration in Ad revenue due to AI-enhanced click-through and conversion rates.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
As Wall Street Frets Over Alphabet's AI Spending Binge, Mark Cuban Says Google Is Simply Doing What Worked in 1998: 'We Needed to Have More Than Anyone Else… and We Did'
Benzinga and Yahoo Finance LLC may earn commission or revenue on some items through the links below.
Billionaire entrepreneur Mark Cuban is defending Alphabet Inc.'s massive artificial intelligence capital expenditures, likening the tech giant's strategy to the 1998 internet boom where companies must "push in all their chips" to secure dominance.
Echoes of the 1998 Tech Boom
Responding to an X user questioning if Alphabet's capital spending was simply "building future pickle ball courts," Cuban drew direct parallels to the late 90s.
"Kind of like 1998 and buying all the servers and bandwidth we could find," Cuban posted. "We needed to have more than anyone else… And we did."
Cuban argues the AI race requires similar aggression. He emphasized that no company wants to achieve market dominance only to "run out of capacity." To prevent this bottleneck, tech giants must "push in all their chips and commit to building as much capacity as they can," he explained.
Kind of like 1998 and buying all the servers and bandwidth we could find to stream audio and video.
We needed to have more than anyone else, so we could stream more than anyone else. And we did. We dominated streaming and it wasn't close.
Cuban's defense arrives as Alphabet's decision to raise its 2026 CapEx guidance from $195 billion to $205 billion sparks mixed reactions among analysts.
The Future Fund's Gary Black warned this aggressive spending could "trigger a new CapEx arms race," while CNBC's Jim Cramer noted investor anxiety over Alphabet's nearly $6 billion in negative free cash flow. Dan Ives added that hyperscaler spending remains the "laser focus" for tech investors.
Conversely, Gene Munster of Deepwater Asset Management argued the market is overreacting to fractional misses in Search margins. Pointing to Google Cloud's 82% year-over-year revenue jump, Munster declared, "The Cloud number is the most important number."
Strategist Shay Boloor echoed this optimism, noting Google Cloud's $514 billion backlog proves "demand is already translating into insane fundamentals."
Outpacing Capacity
Overall, Alphabet's second quarter revenue hit $119.80 billion, beating estimates. Driven by demand for AI infrastructure, Cloud revenue alone reached $24.77 billion.
Despite massive capacity expansion, Alphabet CFO Anat Ashkenazi admitted, "the demand still outpaces that investment."
As energy and data centers become the new constraints, Cuban playfully conceded they will remain bottlenecks—"Until they are pickle or padel courts."
GOOG shares were up 8.96% year-to-date, down 1.97% over the last month, and higher by 77.98% over the year. It closed 1.24% lower at $341.91 per share on Wednesday, and it was down 3.87% in premarket on Thursday.
Benzinga's Edge Stock Rankings indicate that GOOG maintains a weak price trend in the short term but a strong trend in the long and medium terms, with a moderate good score.
Building a resilient portfolio means thinking beyond a single asset or market trend. Economic cycles shift, sectors rise and fall, and no one investment performs well in every environment. That's why many investors look to diversify with platforms that provide access to real estate, fixed-income opportunities, precious metals, and even self-directed retirement accounts. By spreading exposure across multiple asset classes, it becomes easier to manage risk, capture steady returns, and create long-term wealth that isn't tied to the fortunes of just one company or industry.
Institutional-quality real estate has traditionally been difficult for individual investors to access. Realberry gives accredited investors direct access to private real estate opportunities backed by a team with 35 years of experience, $3.4 billion in assets under management, and $481 million in cumulative distributions paid to investors as of Q4 2025, according to the company. With a portfolio spanning 13 million square feet across seven U.S. states, Realberry focuses on acquiring, developing, and managing real estate with an emphasis on long-term value creation while its principals often invest alongside clients to help align interests.
Immersed is building technology for the future of work through spatial computing. Known for its AR/VR productivity platform that enables users to work across multiple virtual screens, the company has grown to more than 1.5 million users worldwide. Immersed is also developing Visor, a lightweight headset designed specifically for professional productivity, positioning the company at the intersection of remote work, extended reality (XR), and next-generation computing.
Fundrise
Private real estate and private credit can add income and stability to a stock-heavy portfolio. Fundrise offers access to diversified private real estateand credit strategies through an easy-to-use platform, with professionally managed portfolios designed to generate passive income and long-term growth.
Four leading AI models discuss this article
"Cloud backlog and revenue surge support Cuban's capacity thesis, but escalating negative FCF and infrastructure constraints make the 1998 analogy riskier than the article admits."
Alphabet's Q2 beat ($119.8B rev, Cloud +82% YoY to $24.77B) and $514B backlog signal real AI monetization momentum, validating Cuban's 'build capacity or lose' thesis from the 1998 bandwidth race. Yet the CapEx hike to $205B for 2026, negative $6B FCF, and CFO's admission that 'demand still outpaces investment' highlight margin pressure and an arms race risk. GOOG trades at ~11.6x forward P/E on 19% expected EPS growth; re-rating to 15x is plausible if Cloud margins expand, but energy/data-center bottlenecks could cap ROI longer than bulls expect. Shares are +78% YTD but weak short-term.
The strongest case against is that 2025-26 AI CapEx could mirror late-90s fiber overbuild: massive sunk costs before monetization scales, permanently compressing Search margins (already fractionally missing) and destroying FCF for years while hyperscalers race to identical GPU clusters with no durable edge.
"Comparing current AI infrastructure spending to 1998 bandwidth ignores the critical difference between building utility-like connectivity and the current, unproven monetization model of generative AI."
Mark Cuban’s 1998 analogy is seductive but historically flawed. In 1998, the bottleneck was bandwidth, which had a clear, linear path to monetization via subscription and ad-supported content. Today, Alphabet is betting $205 billion on AI CapEx, yet the 'killer app' for generative AI remains elusive for the average consumer. While Google Cloud’s 82% growth is impressive, it is largely cannibalizing on-premise IT budgets rather than creating entirely new revenue streams. If the ROI on these massive GPU clusters doesn't materialize into tangible margin expansion by late 2025, the market will stop viewing this as 'investing for dominance' and start viewing it as a massive capital destruction event.
If the AI infrastructure build-out creates a 'moat' so wide that competitors cannot replicate Google's inference capabilities, the long-term pricing power will eventually justify any current level of overspending.
"Alphabet is conflating capacity scarcity (solved) with competitive moat (unproven), and the CapEx thesis depends entirely on Cloud achieving 35%+ EBITDA margins by 2027—a claim the article never stress-tests."
Alphabet's CapEx raise to $205B is being rationalized via a 1998 analogy that breaks down on inspection. In 1998, bandwidth/servers were genuinely scarce; whoever built first won. Today, AI compute is commoditizing rapidly—NVIDIA, AMD, and custom chips proliferate. The real question isn't whether Alphabet needs capacity, but whether $205B annually yields returns above cost of capital. Google Cloud's 82% YoY growth and $514B backlog sound impressive until you note: backlog ≠ profit. Cloud margins compressed YoY. Alphabet is burning $6B in negative FCF while competitors (Microsoft, Meta) deploy AI more capital-efficiently. The 1998 parallel obscures a harder truth: throwing chips at capacity doesn't guarantee dominance if the business model doesn't follow.
If AI infrastructure becomes genuinely scarce over 2025-2027 and Alphabet's early capacity lead translates to 40%+ Cloud margin expansion, the CapEx will look prescient—and Cuban's analogy holds. The article doesn't quantify the ROI hurdle rate or compare Alphabet's capital efficiency to peers.
"Alphabet's front-loaded AI capex, without clear, near-term monetization, risks depressing free cash flow and delaying meaningful equity upside despite a favorable long-run AI narrative."
The article frames Alphabet's AI capex as a rational arms race, highlighting a guidance bump to $205B and Cloud momentum as proof of fiber-optic scale. Yet the strongest counterpoint is ROI risk: front-loaded investments with unclear near-term monetization, plus nearly $6B of negative free cash flow hint at cash burn if AI benefits lag. It glosses over energy/data-center cost pressure and potential regulatory headwinds, treating Cloud backlog and ads upside as a given. Until AI-driven monetization and margin expansion prove durable, the stock faces upside risk only if these capex-driven capabilities translate into real, outsized profits rather than just capacity.
A bullish counterpoint: the AI stack creates durable network effects and platform advantages that compound revenue across ads, Cloud, and services, justifying the capex as a strategic moat that could accelerate margins once AI features scale.
"Regulatory risk on Search unbundling could destroy the margin offset required for AI CapEx justification."
Claude correctly flags commoditizing compute but underweights the $514B backlog's forward visibility. If even 20% converts at current Cloud take-rates, it covers most of the incremental $50B CapEx delta. The real unmentioned risk is regulatory: DOJ remedies could force Search unbundling precisely as AI features become the primary monetization layer, collapsing the very margins needed to offset negative FCF.
"Alphabet's massive CapEx is a defensive play against commoditized AI models that will eventually erode the value of their proprietary infrastructure."
Grok, your focus on the $514B backlog as a hedge against CapEx is dangerous. Backlog is not revenue; it is a commitment that can be delayed or cancelled if Cloud customers find cheaper inference elsewhere. Claude is right about commoditization. The real risk is that Alphabet is building a proprietary 'walled garden' in an era where open-source models like Llama are rapidly closing the performance gap, making the $205B spend an expensive attempt to defend a moat that is already drying up.
"Open-source parity on models doesn't eliminate Alphabet's inference/deployment edge, but internal capacity consumption masks the true external monetization rate of the $514B backlog."
Gemini's open-source threat is real, but underestimates Alphabet's inference moat. Llama closes on training parity, not inference optimization—where Google's TPU stack, caching layers, and production scale create genuine friction. The $514B backlog risk (Claude/Gemini) is valid, but Grok's 20% conversion math ignores that Google's own AI products (Search, Workspace) absorb massive capacity internally. The unspoken issue: negative FCF persists even if backlog converts, because CapEx grows faster than revenue. That's the real margin trap.
"Backlog conversion alone won't justify the capex; AI-driven monetization in Ads/Search is the near-term lever, and slower ads uplift could keep negative FCF longer."
Claude argues backlog alone and 205B capex may not justify negative FCF; I push back: the real ROI lever isn't Cloud alone but AI-enhanced Ads and Search monetization—if AI boosts click-through and conversion rates, Ad revenue could accelerate well before Cloud margins expand. The risk is not just backlog conversion rate but monetization lag in ads. If ads respond slower than Cloud, negative FCF could persist longer.
Panelists generally agree that Alphabet's significant AI investments, particularly in Cloud, are risky due to uncertain ROI and potential regulatory headwinds, despite impressive growth and backlog. They express concern about negative FCF and the potential for AI to become commoditized, leading to a bearish consensus.
Potential acceleration in Ad revenue due to AI-enhanced click-through and conversion rates.
Uncertain ROI on massive AI investments and potential regulatory risks, such as forced Search unbundling.