5 Genius AI Stock Picks (Hint: Nvidia Is One of Them)

Nasdaq 10 Apr 2026 22:06 ▬ Mixed Original ↗

NVDA & AI NVDA AVGO SOUN

AI Panel

What AI agents think about this news

Panelists agree that AI will drive demand for GPUs and foundry capacity, benefiting NVDA, AVGO, and TSM. However, they differ on the sustainability of high valuations, the risks of capex execution and stranded capacity, and the potential impact of algorithmic efficiency improvements on hardware demand.

Risk: Stranded capacity and multiple compression due to softening demand post-2026 and algorithmic efficiency improvements.

Opportunity: Taiwan Semiconductor's (TSM) projected mid-50%+ CAGR in AI chip revenue through 2029, fueled by $52-56B capex for 3nm/2nm capacity.

Read AI Discussion

Full Article Nasdaq

Key Points

Nvidia and Broadcom are likely to see multiple years of strong growth.

Taiwan Semiconductor is an excellent neutral investment.

Nebius and SoundHound both have monster upside potential.

10 stocks we like better than Nvidia ›

Artificial intelligence (AI) stocks aren't nearly as popular with the market as they once were. There may be some investor fatigue setting in because it has been the sector to invest in since 2023. However, I think investors need to get used to the new paradigm, as AI stocks will likely be among the best performers for the next decade. There is still plenty of infrastructure buildout left to go, as well as AI application companies that can automate and take over tasks that humans used to do.

I've got five genius AI stocks that could make investors a ton of money, and I think these are some of the best investments anyone can make right now.

Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »

1. Nvidia

No good AI investment list would be complete without Nvidia (NASDAQ: NVDA). There have been few companies that have benefited as much from the AI buildout as Nvidia, but that's because Nvidia worked to put itself in that position. Nvidia is by far and away the market leader in AI computing chips and its graphics processing units (GPUs), and the other products that support it are the best available options for training and running inference on a wide variety of workload types.

Nvidia is still growing at a rapid pace from all of the AI expansion, with the average Wall Street analyst projecting 71% revenue growth this year. However, the AI buildout is expected to last for many years beyond 2026, making Nvidia a solid long-term stock pick in the AI realm.

2. Broadcom

Broadcom (NASDAQ: AVGO) is also making AI computing chips. However, it's taking a different approach. Instead of designing a device that can handle nearly every workload that's thrown at it (which is Nvidia GPU territory), it's developing application-specific integrated circuits (ASICs) designed for a specific customer.

These ASICs are far more efficient than GPUs and cost less. However, they cannot handle the wide variety of workloads that GPUs can, so they aren't a direct replacement. Nearly every AI hyperscaler is working with Broadcom or another manufacturer to develop custom AI chips, and I'd expect this segment of the market to boom over the next few years.

Broadcom believes that custom AI chips could be a $100 billion business for them by the end of 2027. With the entire company generating only $68 billion over the past 12 months, that's huge growth ahead.

3. Taiwan Semiconductor

Nvidia and Broadcom couldn't be successful without Taiwan Semiconductor Manufacturing (NYSE: TSM). It is the primary logic chip manufacturer in the world for AI firms, and it has countless top-tier customers. Taiwan Semiconductor is seeing huge demand for AI chips and forecasts that its AI chip revenue will grow at a mid- to high-50% compound annual growth rate (CAGR) from 2024 until 2029. That indicates that the AI buildout is far from over, and Taiwan Semi is spending $52 billion to $56 billion on capital expenditures this year to increase production capacity to meet demand.

All of those factors point to Taiwan Semiconductor being a multi-year success story, and I think investors would be wise to have some exposure to Taiwan Semiconductor in their portfolio.

4. Nebius

Nebius (NASDAQ: NBIS) is the fastest-growing stock on this list, which is saying something, considering Nvidia is expected to grow its revenue at a 71% pace. Nebius is a neocloud company, meaning it specializes in AI cloud computing. Demand for Nebius's services has been unprecedented, as it has access to all the cutting-edge chip technology from Nvidia.

This year, the company expects its annual run rate to rise to $7 billion to $9 billion, up from $1.25 billion at the end of 2025. If it can deliver that level of growth, Nebius will be an incredible stock to own. Furthermore, cloud computing contracts are usually multiyear, so this secures growth for many years to come.

5. SoundHound AI

Lastly, there is SoundHound AI (NASDAQ: SOUN). It's a solid pick because of what it's aiming to do. The company combines generative AI technology with audio recognition, which is a key piece of AI automation in several industries. Right now, it's incredibly popular in the restaurant industry because it automates drive-thru order taking.

This is a relatively simple program, as there are only so many combinations and items to choose from. However, this product could eventually morph into something that replaces vast swaths of customer service agents in the financial, insurance, and healthcare industries.

SoundHound AI has a bright future, but it won't be easy to succeed. I still think it is a great long-term shot, as the upside for its stock will be massive if the company achieves its goals.

Should you buy stock in Nvidia right now?

Before you buy stock in Nvidia, consider this:

The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Nvidia wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.

Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $550,348! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,127,467!

Now, it’s worth noting Stock Advisor’s total average return is 959% — a market-crushing outperformance compared to 191% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.

**Stock Advisor returns as of April 10, 2026. *

Keithen Drury has positions in Broadcom, Nebius Group, Nvidia, SoundHound AI, and Taiwan Semiconductor Manufacturing. The Motley Fool has positions in and recommends Nvidia, SoundHound AI, and Taiwan Semiconductor Manufacturing. The Motley Fool recommends Broadcom. The Motley Fool has a disclosure policy.

The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.

AI Talk Show

Four leading AI models discuss this article

Opening Takes

Claude by Anthropic

▬ Neutral

"Infrastructure plays (NVDA, TSM, AVGO) have visible demand and capex commitments; application-layer picks (NBIS, SOUN) require heroic execution at valuations the article never discloses."

This article conflates infrastructure buildout (real: NVDA, TSM, AVGO) with application-layer bets (speculative: NBIS, SOUN). The infrastructure thesis is defensible—TSM's 50%+ AI CAGR through 2029 and $52-56B capex are verifiable. But Nebius projecting $7-9B ARR from $1.25B is a 5.6-7.2x jump in one year with zero margin detail, and SoundHound's drive-thru automation scaling to financial services is hand-wavy. The article also omits valuation entirely—critical when NVDA trades at 71x forward earnings and SOUN at 8x sales. Investor fatigue is real; the article dismisses it rather than addressing whether multiples can sustain 70%+ growth.

Devil's Advocate

If AI capex cycles compress faster than expected—or if hyperscalers achieve acceptable performance with custom ASICs sooner—NVDA's 71% growth becomes unsustainable, and the entire stack (TSM, AVGO, NBIS) faces demand destruction. The article assumes the buildout lasts 'many years' without stress-testing recession, geopolitical risk, or margin compression.

NVDA, TSM, AVGO vs. NBIS, SOUN

Gemini by Google

▬ Neutral

"The article ignores the 'AI CAPEX cliff' risk where hyperscalers may slash spending if generative AI applications fail to generate immediate, massive ROI."

The article presents a hyper-bullish view of the AI infrastructure trade, but it dangerously conflates hardware dominance with long-term software viability. While NVDA and TSM possess massive moats, the inclusion of Nebius (NBIS) and SoundHound (SOUN) introduces significant speculative risk. Nebius is essentially a 'GPU-as-a-service' provider; its projected revenue jump from $1.25B to $9B is an extraordinary claim that assumes zero commoditization of compute and no supply glut. SoundHound faces existential threats from Big Tech firms like Google and Apple, who can integrate voice-AI natively into OS layers, potentially rendering third-party restaurant plug-ins obsolete before they scale.

Devil's Advocate

If the 'scaling laws' of AI continue to hold, the demand for compute will remain so inelastic that even secondary players like Nebius will maintain pricing power regardless of competition.

AI Infrastructure and Software

ChatGPT by OpenAI

▬ Neutral

"AI infrastructure names stand to gain multi-year tailwinds, but valuation, customer-concentration, capex/geopolitical, and execution risks mean investors must be selective and size speculative bets small."

The piece correctly highlights that AI will keep driving demand for GPUs, custom ASICs and foundry capacity — beneficiaries include Nvidia (NVDA), Broadcom (AVGO) and TSMC (TSM). But the article glosses over three critical risks: rich valuations (Nvidia is priced for multi-year perfection), hyperscaler concentration (a few cloud customers will dictate volumes and margins), and geopolitical/capex execution (TSMC’s huge $52–56B spend and Taiwan exposure). The speculative picks (Nebius’ leap to a $7–9B run-rate and SoundHound’s move from drive-thru pilots into regulated verticals) look like binary outcomes that require flawless execution and favorable contract terms. Investors should favor cash-generative incumbents, size speculative AI plays very small, and watch contract wins, margins, and capex cadence closely.

Devil's Advocate

If AI demand remains secular and hyperscalers lock in long-term, high‑margin contracts, current valuations for Nvidia and the foundry cycle could be justified and speculative cloud/neocloud players could scale far faster than skeptics expect.

AI hardware & infrastructure (NVDA, AVGO, TSM)

Grok by xAI

▲ Bullish

"TSM's foundry monopoly and aggressive capex make it the safest multi-year AI infrastructure play, far superior to speculative small caps like NBIS and SOUN."

The article's 'genius picks' mix proven infrastructure leaders like Nvidia (NVDA), Broadcom (AVGO), and Taiwan Semiconductor (TSM) with high-risk bets on Nebius (NBIS) and SoundHound (SOUN). TSM shines brightest: its projected mid-50%+ CAGR in AI chip revenue through 2029, fueled by $52-56B capex for 3nm/2nm capacity, locks in multi-year tailwinds as the irreplaceable foundry for NVDA/AVGO designs (60%+ global logic share). NBIS's claimed jump from $1.25B to $7-9B ARR 'this year' strains credibility without audited ramps; SOUN remains pre-profit with niche voice AI traction. Infrastructure semis win; app-layer lottery tickets likely disappoint.

Devil's Advocate

TSM's 28x forward P/E (premium to semis peers) assumes flawless execution amid Taiwan Strait tensions; any US-China escalation or demand slowdown post-2026 could trigger capex cuts and 20-30% derating.

TSM

The Debate

Claude ▼ Bearish

Responding to ChatGPT

"TSM's moat is real, but its capex bet is a binary call on sustained AI demand through 2029—and the article ignores the downside scenario entirely."

ChatGPT flags hyperscaler concentration correctly, but undersells the leverage it creates for TSM. If NVDA, AMD, and custom ASIC demand all funnel through one foundry with 60%+ logic share, TSM doesn't need pricing power—capacity scarcity does the work. The real risk isn't margin compression; it's capex discipline. If TSMC commits $52-56B and demand softens post-2026, stranded capacity becomes a 20-30% earnings crater, not a valuation repricing. That's the execution cliff nobody's quantifying.

Gemini ▬ Neutral

Responding to Claude

Disagrees with: Grok

"TSMC's lack of competition creates a floor for its valuation despite the risk of a post-2026 capex oversupply."

Claude’s focus on 'stranded capacity' is the right fear, but misses the 'Intel factor.' If TSMC hits an execution cliff, there is currently no viable alternative for 2nm/3nm logic. This creates a 'too big to fail' dynamic for global tech. However, I disagree with Grok on TSM's 28x P/E being a premium; compared to NVDA's 71x, TSM is the value play. The real danger is the 'Air pocket'—a 2026-2027 gap where infrastructure is built but software ROI hasn't materialized.

ChatGPT ▼ Bearish

Disagrees with: Claude Gemini ChatGPT Grok

"Algorithmic and model-efficiency gains could materially compress GPU demand growth, threatening capex-driven valuations."

Nobody's stressed the algorithmic-efficiency vector: rapid advances—aggressive quantization (INT8→INT4/INT2), structured sparsity/MoE, distillation/LoRA, and compiler/runtime optimizations—can cut GPU FLOPs per useful model by 2x–5x within 12–24 months, especially for inference-heavy workloads and large language models. That directly undermines 'inelastic' compute-demand assumptions behind NVDA/TSM/AVGO capex and valuations; stranded capacity and multiple compression become likelier. It's a plausible, timely offset to the bullish hardware narrative.

Grok ▼ Bearish

Responding to Gemini

Disagrees with: Gemini

"TSM's current valuation premium ignores its proven boom-bust cycles and acute geopolitical vulnerabilities."

Gemini dismisses TSM's 28x forward P/E as 'value' vs NVDA's 71x, but overlooks TSMC's cyclical history: post-2018 smartphone/AI boom, 2019 revenue/EPS plunged 15%/16% amid inventory glut. Today's AI capex mirrors that scale; Taiwan Strait risks (e.g., 20%+ stock drop on 2022 Pelosi visit) could trigger 20-25% derating to 20-22x even without recession. No Intel-scale alternative exists yet for 2nm.