Key Moments
The IPO Comeback: Why Tech Giants Are Finally Going Public | All-In Liquidity IPO Panel
All-In PodcastWant to know something specific about what's covered?
We've already dissected every moment. Ask and we will deliver (with timestamps).
Key Moments
Tech IPOs are making a comeback, with companies like Cerebras and Planet Labs highlighting the renewed investor interest in public markets. While going public brings liquidity and validation, the core business operations remain unchanged.
Key Insights
2026 is predicted to be an all-time record year for IPOs, with AI companies leading the charge.
Andrew Feldman (Cerebras) noted that his company's decade-long struggle to go public paradoxically helped them time the market effectively.
Will Marshall (Planet Labs) stated that launch costs for satellites have decreased by about 4-5x over the last 10 years, making space ventures more accessible.
By approximately 2027-2028, it is projected to be cheaper to deploy data centers in space than on Earth, with solar power offering a significant advantage.
Andrew Feldman explained that AI enabled computers to successfully address problems they were previously poor at, such as image and language processing, thereby opening vast new markets.
Historically, studies show more money is made after a company goes public than before, emphasizing the importance of long-term investment post-IPO.
The IPO market is experiencing a resurgence
The year 2026 is shaping up to be a record-breaking year for Initial Public Offerings (IPOs), with AI companies at the forefront of this resurgence. Companies like Cerebras, focused on AI silicon, and Planet Labs, specializing in space-based Earth imaging and data, are highlighted as prime examples of this trend. Despite the excitement and increased capital that comes with going public, founders Andrew Feldman of Cerebras and Will Marshall of Planet Labs suggest that the fundamental operational aspects of their businesses remain largely unchanged. The immediate impact is often centered on employee morale and providing liquidity for early investors and employees, rather than an overnight transformation of business operations. Feldman specifically mentioned that the extensive work and meetings involved in the IPO process, while necessary, do not inherently advance engineering projects or solve existing supply chain issues.
Navigating the IPO journey: Challenges and employee impact
Andrew Feldman shared that his company, Cerebras, faced a challenging decade-long path to going public, which ultimately contributed to their well-timed market entry. He emphasized the significant event it was for long-term employees and their families, fostering a sense of pride and external validation. Despite the celebratory aspects, Feldman stressed a swift return to core work, stating, "Now what? Back to work." This sentiment underscores the continuous operational focus required post-IPO. The process itself was described as arduous, involving numerous meetings and document reviews with seemingly little direct value added to the core business. However, the influx of capital and increased profile can offer advantages in selling to enterprise customers and securing new supply or vendor relationships, should they have been poor prior to the IPO.
Planet Labs: From niche data to global impact in space
Will Marshall discussed Planet Labs' journey as one of the early space stocks, noting that the market is increasingly understanding the transformative power of space data. Planet Labs operates a fleet of approximately 200 satellites that image the entire Earth daily, providing up-to-date satellite imagery that can be used for time-series analysis. This data is crucial for diverse sectors including agriculture, energy, civil governments (for disaster monitoring like floods and fires), and security applications. Marshall highlighted that AI is significantly lowering the barrier to entry for accessing and utilizing this vast amount of data. While security applications are a growing part of their business, particularly due to current geopolitical demands, Marshall clarified that Planet Labs serves a broader range of clients, including farmers, energy companies, and NASA, positioning themselves beyond just a military-focused entity. The company's legitimacy and long-term viability are enhanced by its public company status, assuring large government clients who rely on their continuous data streams.
The economics of launching infrastructure into space
Marshall cited that satellite launch costs have decreased significantly, by about 4-5 times over the last decade. A more crucial, yet less known, development is the miniaturization of satellites. Today's satellites, weighing tens or hundreds of kilograms, can perform tasks equivalent to or exceeding those of older, 20-ton satellites that cost billions. This revolution, akin to the mainframe to desktop computer shift, is unlocking numerous new applications. Looking ahead, Marshall anticipates that by the time launch costs fall to around $200-$300 per kilogram (from the current ~$1,000/kg), it will be economically more viable to place data centers in space. He projects this could happen in the next 2-3 years, potentially driven by technologies like Starship. In space, solar panels can achieve five times more energy generation by remaining in a sun-synchronous orbit, eliminating the need for batteries and reducing infrastructure complexity to solar panels, chips, and RF communications.
The rebirth of silicon driven by AI's expanding problem domain
Andrew Feldman explained that AI has fundamentally changed computing by enabling machines to effectively address problem domains previously considered intractable. For most of computing history, computers excelled at numerical calculations but struggled with unstructured data like images and natural language. AI, emerging significantly around 2015-2016, opened the door for computers to process and derive insights from images, and to generate and understand language beyond mere storage. This expansion, coupled with the increasing collection of data (like Planet Labs' Earth imagery), created a massive demand for compute power. Feldman's company, Cerebras, took a contrarian approach by betting on dedicated silicon rather than merely improving existing architectures like GPUs. Their strategy involved creating a very large, 'dinner plate' sized chip with memory integrated directly alongside compute. This architecture allows for significantly faster data movement, resulting in performance gains, such as being 15-18 times faster than GPUs for certain AI tasks. Feldman stressed the critical need for real-time AI responses, comparing the market for slow search or dial-up to be non-existent, thus necessitating immediate deliverability, which their architecture aims to provide.
Rethinking liquidity and long-term value post-IPO
The panel discussed the journey of investors as companies go public. Will Marshall noted that for Planet Labs, which went public in 2021 via SPAC, much of the value creation (90%) occurred in the three to four years following the IPO, with most early investors wisely holding onto their shares. This contrasts with a historical tendency where LPs might pressure for immediate share distribution. Andrew Feldman echoed this sentiment, stating that "historically more money's made after IPO than before." He highlighted that while venture capital allows for only modest investments in early-stage companies, the post-IPO phase offers greater opportunities for substantial absolute returns, especially for larger investors. Both founders suggested that companies going public sooner, even at lower valuations, could have enabled public market investors to capture some of the subsequent growth, as seen with Planet Labs. Innovations like 'dribble lockups,' used by Cerebras and potentially SpaceX, allow shares to be gradually released over time based on performance hurdles, offering a more flexible approach to post-IPO liquidity for investors.
The synergistic future of AI and space
Will Marshall articulated a vision where AI and space converge to create 'planetary intelligence.' He argued that current large language models (LLMs), while powerful, are primarily trained on text and lack understanding of the real world. By integrating real-world data from space-based sensing systems – like farm field conditions, flood status, or security situations – into AI models, previously unmet needs can be addressed. This fusion would unlock gazillions of new applications, moving beyond 'large language models' to 'large earth models' or 'planetary intelligence.' This integration promises to enable AI to solve real-world problems, driving significant economic growth and creating novel 'planetary compute systems in space.' While the exact timeline is debated, the convergence of AI and space is seen as an inevitable and transformative force for the next few years, creating a substantial new economy.
Mentioned in This Episode
●Products
●Companies
●Organizations
●People Referenced
Space Launch Costs vs. Projected Data Center Viability
Data extracted from this episode
| Launch Cost per Kilogram | Projected Viability Point | Current Launch Cost |
|---|---|---|
| $200 - $300 | Makes space data centers cheaper than ground | $1,000+ |
Cerebras Chip Performance vs. GPU
Data extracted from this episode
| Chip Type | Performance Multiplier | Benefit |
|---|---|---|
| Cerebras Chip (with large die and adjacent memory) | 15-18x faster | Faster answers, more enjoyable engagement, ability to solve harder problems |
Common Questions
Going public involves a significant amount of 'garbage' like excessive meetings and document reviews without direct business progress. While it brings more capital and profile, core business operations and vendor relationships remain largely unchanged.
Topics
Mentioned in this video
A company focused on AI silicon, with its founder and CEO Andrew Feldman discussing their experience going public.
A company that images the entire Earth daily with a fleet of satellites, with co-founder and CEO Will Marshall discussing its public market journey, the role of space data and AI, and its applications in agriculture, energy, and civil governments.
Partnering with Planet Labs to launch TPUs into space and conducted a study on space-based data centers potentially years prior.
Tech investment firm led by Brad Gersner, mentioned in the context of investing and IPOs.
Mentioned as part of satellite constellations transporting data around the planet.
Mentioned for its role in reducing rocket launch costs, its Starship program, and its potential future IPO impacting valuations.
A major player in AI silicon and GPUs, mentioned as a competitor and a company whose architecture Cerebras aims to surpass.
Mentioned as a competitor in the silicon market, whose architecture is contrasted with Cerebras's approach.
A prominent AI research lab that uses Cerebras's chips, experiencing significant speed improvements.
Venture capital firm, mentioned as an investor in Planet Labs.
Investor in Cerebras, mentioned in the context of lockup periods and their potential distribution of shares.
Mentioned as one of the 'mega IPOs' or companies with significant private market value accrual, alongside OpenAI and SpaceX.
Founder and CEO of Cerebras Systems, discussing the challenges and realities of going public, the excitement of employees, and the importance of not changing core business operations.
Co-founder and CEO of Planet Labs, discussing the synergy between space and AI, the company's approach to earth imaging, and the potential for space-based data centers.
Founder and CEO of Altimeter Capital, a leading tech investment firm, who is introduced as a panelist.
Mentioned in an anecdote by Jason Calacanis about a badge at Davos.
Mentioned as an early investor in Planet Labs.
Mentioned as a founder of Founders Fund, an investor in Planet Labs.
Mentioned as an investor in Planet Labs through DST.
Mentioned regarding SpaceX's ambitions and potential valuations, and his view on Starship's launch cost reduction timeline.
More from All-In Podcast
View all 416 summaries
32 minDan Loeb: The Lost Art of Short Selling, and Why Stock Picking is Back
33 minThomas Laffont: The $4T AI IPO Wave Is Coming… and We’ve Never Seen Anything Like It
30 minBill Ackman: Investment Strategy, What the Market is Missing, How AI Breaks Businesses
33 minOpenAI CFO Sarah Friar on IPO, AI Rivalries, New Device, and Spending $100B+ on Compute
Ask anything from this episode.
Save it, chat with it, and connect it to Claude or ChatGPT. Get cited answers from the actual content — and build your own knowledge base of every podcast and video you care about.
Get Started Free