All-In Podcast

The IPO Comeback: Why Tech Giants Are Finally Going Public | All-In Liquidity IPO Panel

32 minJun 6, 2026
Key Themes
IPO dynamicspublic market accessAI siliconsatellite imageryspace computefrontier techliquidity and lockupspost-IPO growth
Summary

A wide-ranging IPO panel on public listings, AI infrastructure, and the next frontier in space data

This episode brings together leaders from Cerebras and Planet Labs for a conversation about what it means to go public, how IPOs affect companies and people, and why public markets may be reopening to frontier tech. The discussion moves from employee morale, liquidity, and customer credibility to larger themes in AI silicon, satellite imagery, and the possibility of space-based compute. Across both chapters, the panel argues that many of the most meaningful gains in breakthrough companies happen after the IPO, not before.

1
An IPO changes status more than operations.

A recurring point in the episode is that going public adds reporting, scrutiny, and process, but the core work of building the company stays largely the same. That makes the IPO less of a finish line and more of a transition in how the business is perceived and financed.

2
Public-company status can create practical and emotional value.

The speakers emphasize liquidity, cash access, credibility, and customer confidence, but also the pride employees and families feel when a long effort ends in a public listing. The episode treats these as real benefits beyond the stock price itself.

3
AI is making satellite data easier to use.

Planet Labs’ business is presented as vast but inherently complex data infrastructure. The panel argues that AI lowers the barrier to entry for customers, making earth-observation data more usable across agriculture, energy, government, and security applications.

4
Space infrastructure becomes more plausible as launch costs fall.

The episode’s space-compute section hinges on the idea that better economics can unlock entirely new architectures. Cheaper launch, satellite miniaturization, and abundant solar power are presented as the ingredients that could make orbital compute and other space-native services more practical.

+1 more
5
Specialized hardware can outperform general-purpose approaches in the right workload.

Cerebras is used to argue that when data movement is the bottleneck, architecture matters as much as raw compute. The discussion frames dedicated silicon as a way to deliver materially better performance for certain AI workloads than more general-purpose GPU-based systems.

6
The IPO is not the end of value creation.

The panel repeatedly argues that frontier companies often generate a large share of their ultimate value after the listing, not before it. That view reframes IPO timing as a liquidity and ownership question rather than a signal that the highest-growth phase is over.

7
Frontier companies may be returning to public markets sooner.

A major thread in the episode is that public markets may once again be willing to underwrite long-duration innovation stories earlier than before. That could reshape how founders think about staying private, who captures the upside, and how much discipline public scrutiny adds.

Select any chapter text to Deep Dive with AI
01IPO reflections from Cerebras and Planet Labs CEOs

A roundtable with Andrew Feldman and Will Marshall on the lived experience of going public, the limited operational change that follows an IPO, and the benefits public status can bring for employees, customers, and long-term company legitimacy. The conversation also connects Planet Labs’ Earth-imaging business to broader shifts in AI, satellite access, and cheaper launch economics.

Going public adds process and scrutiny, but it does not fundamentally change the day-to-day core of the business.
Cerebras’ IPO is described as unusually difficult, but the listing carried symbolic value for employees and families.
Planet Labs shows how public-market understanding can improve over time as the business becomes better known.
Public-company status can improve liquidity, capital access, credibility, and customer confidence.
Planet Labs’ business centers on daily Earth imaging and time-series analysis for sectors like agriculture, energy, civil government, security, and defense.
AI and falling launch costs are making satellite data more accessible and expanding the space sector.
02Space data centers, AI silicon, and liquidity after IPO

The panel expands into a vision of space-based data centers, arguing that cheaper launch and abundant solar power could eventually make compute in orbit economically attractive. It then turns to Cerebras’ chip architecture and the case for specialized silicon, before closing with a discussion of IPO timing, lockups, and how much value in frontier companies may still be created after listing.

Space-based compute is framed as a natural extension of satellite infrastructure and AI data processing.
The near-term opportunity remains Earth observation and AI analysis of imagery.
Cheaper launch and continuous solar power are presented as the core economic logic for orbital data centers.
Cerebras’ architecture is described as minimizing data movement by keeping memory close to compute.
The panel contrasts dedicated silicon with general-purpose GPU approaches.
The discussion argues that much of the upside in frontier companies can happen after the IPO.
Public markets may be reopening sooner to frontier companies instead of forcing them to stay private longer.