All-In Summit 2025: Hangin with the Besties
Animation by Sora
Only at All-In Summit can you see so many tech luminaries in one place. This year’s Summit, again set in La La Land, brought together an incredible array of thought leaders – from provocative to deeply intellectual, and everything in the spectrum between.
We saw Joe Tsai of Alibaba (a thoughtful bridge-builder), Rene Haas of Arm (chip magnate), Roelof Botha of Sequoia (a real outlier in venture land), Demis Hassabis of DeepMind (Nobel Prize winner), Elon Musk of Tesla and SpaceX (now only the world’s second richest man), Mark Cuban (a real human), Orlando Bravo of Thoma Bravo (best of the best in Tech PE), Alex Karp of Palantir (agent provocateur – “don’t believe a word I say!”), Dara Khosrowshahi of Uber (among the great CEOs), U.S. Energy Secretary Chris Wright (the gas man), Vlad Tenev of Robinhood (democratising wealth), Keller Rinaudo Cliffton of Zipline (saving lives in Rwanda), Cathie Wood of ARK Invest (either completely wrong or will be very, very rich), Eric Schmidt former Chair of Google (another luminary), Anatoly Yakovenko of Solana (crypto bro), Kyle Samani of Multicoin Capital (crypto bro 2), and Adena Friedman of Nasdaq (corporate meets web3).
Each speaker brought their perspective on where technology is heading and where the value will be created. But what’s interesting is who wasn’t there, and what wasn’t said. The Zuck wasn’t to be seen, nor were Nvidia, Microsoft or Amazon. There wasn’t last year’s MAGA hype (thank goodness) and I thought the Besties were a little less buoyant. One or two topical issues (like the Federal Government taking 10% of Intel) were avoided. And Elon seemed more focused – no dirty jokes, and certainly no politics (other than proclaiming “the Government is unfixable!”)
Still, you cannot squeeze everything into two days and you cannot control what such a diverse group of leaders will say - my synopsis is the Besties did a damned good job of it. Here’s our breakdown, without the political stuff.
Article continues below.
Like this article? Join the thousands of tech founders, board members and investors who subscribe to our free monthly newsletter, Tech Round-Up. Sign up below!
AI: From Hype to Infrastructure
At All-In Summit 2024, the AI conversation was dominated by speculation. Could AGI be close? Would the wave of investment in large language models generate sustainable returns? Investors were cautious, concerned about commoditization of foundation models and the limited number of clear monetization paths.
A year on, the tone has shifted dramatically. In 2025, AI is no longer treated as a speculative layer of apps but as core infrastructure. Leaders from DeepMind, Tesla, and Eric Schmidt stressed that AI is the ultimate accelerator for scientific discovery, engineering, and productivity. The focus has moved from asking if AI will matter to where profits will accrue: inference at the edge, robotics integration, and domain-specific applications in healthcare, logistics, and defense.
The consensus now is that edge AI - billions of devices and robots making local decisions - will dwarf centralized training workloads. The opportunity lies in the platforms that make this possible: chips, data centres, multimodal models, and robotics ecosystems. The investor mindset has matured from speculative enthusiasm to platform conviction.
Chips: The Bottleneck Becomes the Battleground
In 2024, Nvidia’s dominance in GPUs was the headline. Scarcity of chips was the primary concern, and investors wondered whether challengers could ever make a dent. The conversation was about supply chains, not strategy.
By 2025, the discussion has bifurcated. Nvidia remains dominant in training workloads, but Arm’s CPUs are emerging as the backbone of inference. Leaders forecast two distinct markets: a small number of massive, complex chips for training and swarms of smaller, distributed chips for inference. Robotics alone could create chip demand that dwarfs the smartphone era.
The strategic angle has intensified. Semiconductor manufacturing capacity, lithography equipment, and rare-earth processing are now matters of national power. Western governments are urged to coordinate investments to avoid ceding ground to Asian fabs. Investors today see chips not just as a cyclical hardware story but as the backbone of the AI century.
Robotics and Autonomy: The Physical Internet
In 2024, robotics was treated with skepticism. Musk teased Optimus, but most dismissed it as aspirational. Drone delivery was still seen as futuristic, and most investors placed robotics as a “maybe” play in the distant future.
A year later, robotics is at the center of strategic conversation. Musk now describes Optimus as potentially “the most valuable product in history” (he has certainly been known to get things wrong, but who can deny his achievements?) DeepMind’s progress in generative physical environments and Schmidt’s warnings on algorithmic warfare reinforce robotics as an inevitability. Meanwhile, Zipline has shown what scaled autonomy looks like in practice: millions of healthcare and retail deliveries, with drones proving faster, cheaper, and cleaner than traditional couriers.
The narrative has shifted from curiosity to inevitability. Robotics is now framed as the “physical internet” - a system of autonomous platforms moving goods and performing labor at scale. The investor opportunity is clear: companies that combine hardware, AI, and regulatory defensibility will dominate.
Energy: Powering the AI Economy
Energy in 2024 was framed mostly through ESG and climate politics. Tech insiders acknowledged rising data centre demand, but the issue felt secondary to debates over subsidies and carbon targets.
By 2025, energy is central. With data centres demanding dozens of gigawatts of new capacity, the US Energy Secretary underscored the reality: hydrocarbons still supply 85% of global energy. Natural gas is scaling fastest, nuclear remains unmatched in density but stifled by regulation, and by an almost apocalyptic fear factor. Renewables are constrained by intermittency and play a surprisingly small role on the global energy stage.
The conversation has shifted from ideology to pragmatism. The AI economy will be power hungry, and profits will accrue to those who provide reliable, affordable electrons. For investors, the opportunity is in flexible natural gas assets, deregulated nuclear plays (if politics allow), and distributed solar-plus-battery infrastructure.
Market Structure: Public vs Private Capital
Throughout both years, the structural decline of listed companies (with company numbers down by 40% over 25 years on American exchanges) has been accompanied by the rise (and rise) of private equity, reflecting regulatory burden and the appeal of private markets. Private equity funds like Thoma Bravo can raise tens of billions and deliver predictable returns, which institutional LPs prize. In 2024 this was a warning sign; in 2025 it is accepted as the new normal.
By contrast, venture capital faces diminishing returns. Sequoia noted in 2025 that $150–200B goes into VC annually, but exits cannot mathematically keep pace. Venture firms must focus on rare outliers and long-term compounding, while sovereign funds and direct investors reshape late-stage dynamics. For investors, the takeaway is that PE is on a roll, while VC is underperforming, there are too many underperforming GPs and the sector faces structural headwinds.
Finance and Crypto: From Distrust to Internet-Native Capital Markets
In 2024, crypto was on the defensive. Scandals and regulatory uncertainty left investors skeptical. The debate centered on survival and whether tokens had any use beyond speculation.
By 2025, the mood is transformed. With the GENIUS Act passed and the Clarity Act pending, the US has embraced crypto as infrastructure. Leaders from Solana, Multicoin Capital and Robinhood talked openly about on-chain capital markets replacing legacy systems. Robinhood is positioning to capture the $130 trillion wealth transfer with a super-app that blends deposits, trading, and tokenized securities. Solana touts sub-200ms execution, 1000x faster than Ethereum. Kyle Samani predicts legacy markets will be “consumed, not adapted.”
If you listen to the crypto bros, investor focus has shifted from whether crypto has legitimacy to how quickly it can absorb traditional finance. Infrastructure—blockchains, exchanges, middleware—and applications that embed trading in social contexts are now live opportunities. Crypto has moved from distrust to inevitability. Or so the crypto bros think.
Capital Allocation and Trust
The capital cycle discussion also matured between 2024 and 2025. In 2024, firms spoke of resetting expectations after the 2021 bubble. By 2025, Sequoia sharpened its point: outliers drive returns, not averages, and holding winners through IPOs is the alpha strategy. Palantir underscored another side of defensibility: trust architecture. Immutable logs and permissioning, once dismissed as cumbersome, are now seen as critical for enterprise AI adoption. (Alex Carp from Palantir had one helluva lot more to say, but not that much was about technology…)
The theme has shifted from recovering after excess to building moats in a world awash with capital. Discipline, trust, and defensibility are the new investor mantras.
Healthcare and Transparency: A Systemic Disruption
Healthcare disruption was a side theme in 2024. AI in drug discovery or diagnostics was promising but seen as long-term.
In 2025, Mark Cuban has put healthcare at the center of the tech disruption narrative. CostPlus proves that lean, AI-enabled companies can attack entrenched inefficiencies with minimal headcount. Radical transparency in pricing and direct-to-provider networks are positioned as systemic shifts. The $4 trillion US healthcare market is now firmly in investors’ sights as one of the last great inefficiencies ripe for technology-driven disruption.
Demis Hassabis of DeepMind was very clear about the imminent breakthroughs in health technology in an AI age.
Noah’s Ark
Cathie Wood from Ark Invest doubled down on her trademark optimism—forecasting 7%+ GDP growth, vanishing inflation, and a world reshaped by converging platforms like AI, robotics, energy storage, blockchain, and genomics. She argues these forces could drive 40–50% annual returns for disruptive innovation, finally broadening markets beyond the Mag 6. On the bold calls, she put Bitcoin at $1.5M, Tesla at $2,600 fueled by robotaxis rather than cars, and tipped healthcare AI as the most underpriced opportunity in markets today. Whether you see her as a visionary or just heavily caffeinated, Wood made disruption sound not only inevitable but wildly profitable—and she did it with her usual flair for turning spreadsheets into sermons.
The Investor Takeaway
In 2024 across AI, chips, robotics, energy, crypto and capital markets, investors were asking whether these shifts were real or investable. By 2025, the conversation had become about how to capture the value. The mood has shifted from tentative speculation to conviction. For founders, this is the decade to build platforms with global defensibility. For investors, the winners will be those who back enabling infrastructure and rare outliers with staying power. Disruption is no longer in question; the only question is who captures the profit pools as industries converge into the digital-first economy.
My Own Takeaway
Attending All-In Summit is like drinking from a fire hose. There’s an enormous amount of content to digest, from many of the world’s smartest techies. You need a big filter and an even bigger bullshit detector. And all the while, you need to remind yourself that you’re in a $7,500-per-ticket echo chamber.
I attend to get my annual Tech tune-up (a mix of inspiration, wonderment and terror), to escape from the island mentality that pervades Down Under, and to learn. Mission accomplished. By the end of the Summit, Down Under never seemed so good.
Roger Sharp
Founder of North Ridge Partners