Skip to content
WhySoGeek.
News

Bezos Backs Flourish's $500M Brain-Inspired AI

Flourish raised $500M with Jeff Bezos backing to build brain-inspired AI aiming for laptop-level power draw instead of server racks.

Sam Carter 7 min read
Cover image for Bezos Backs Flourish's $500M Brain-Inspired AI
Photo: brewbooks / flickr (BY-SA 2.0)

Most AI startups promise bigger models. Flourish is promising a smaller power bill. The company raised $500 million, with Jeff Bezos among its backers, to build brain-inspired AI that aims to run on the power of a laptop rather than a server rack.

Quick answer

Flourish, a startup building brain-inspired AI, raised $500 million at about a $2.5 billion valuation in a round that closed around June 4, 2026. Backers include Jeff Bezos, Lux Capital, and GV. Flourish is building a system it calls Cortex AI that studies real neurons and their connections, aiming to run in the 20 to 50 watt range, an order of magnitude below conventional AI hardware.

Key takeaways

  • Flourish raised $500 million at roughly a $2.5 billion valuation.
  • Jeff Bezos is a backer, alongside Lux Capital and GV.
  • The approach is connectomics, mapping real neurons and their connections.
  • The power target is 20 to 50 watts, laptop-level rather than server-rack.
  • The founders are notable, including Internet Explorer creator Thomas Reardon.

Betting against brute force

The dominant AI strategy is to scale up: more parameters, more data, more graphics processors, more electricity. Flourish is betting the opposite direction. Its pitch is to stop approximating how the brain works and instead study it directly, mapping real neurons and their connections in a field called connectomics, then building AI on those principles.

The payoff, if it works, is efficiency. Flourish's system, Cortex AI, targets a power draw of 20 to 50 watts, roughly a laptop, rather than the kilowatts a server rack consumes. That would be an order-of-magnitude improvement over conventional AI hardware and would directly address the industry's growing energy problem.

ApproachStrategyEnergy profile
Mainstream AIScale up models and computeKilowatts, data-center racks
Flourish / Cortex AIStudy and emulate real brains20 to 50 watts, laptop-level
An illustration of neurons connected in a network resembling brain tissue
Photo: National Institutes of Health (NIH) / flickr (BY-NC 2.0)

Who is behind it

The founders give the effort credibility. Flourish was co-founded by Thomas Reardon and Rob Williams. Reardon created Internet Explorer at Microsoft and later founded CTRL-labs, a brain-computer interface company that Meta acquired in 2019. Williams is a former Amazon senior executive. That pedigree in both software and neuroscience-adjacent hardware is a large part of why serious investors showed up.

Jeff Bezos initially committed roughly $50 million and nearly doubled his stake after other high-profile investors joined, according to reports. The round also drew Lux Capital and GV, Alphabet's venture arm. The valuation landed around $2.5 billion for a company still deep in research.

Why the power angle matters

The energy cost of AI is no longer a footnote. Training and running large models consumes enormous amounts of electricity, straining grids and data-center budgets, and the trend is upward. Any credible path to running capable AI at a fraction of the power is strategically valuable, which is why a research-stage company can raise $500 million on the promise.

The caution is that brain-inspired computing has a long history of ambitious claims and slow progress. Connectomics is genuinely hard, and translating neuroscience into working, efficient AI hardware is unproven at scale. This is a long-horizon bet, not a product you will buy next year. For the scale of the problem Flourish is targeting, our coverage of AI data center energy demand lays out why efficiency has become urgent.

FactorBull caseBear case
Energy efficiencyOrder-of-magnitude gains possibleUnproven at scale
Founder pedigreeDeep software and neuro experienceTrack record does not guarantee results
ConnectomicsNovel, principled approachHistorically slow to deliver
TimelineLong-term strategic payoffYears from a product

What this means

For the AI industry, Flourish represents a meaningful counter-bet to the scale-everything consensus. If even part of its efficiency thesis pans out, it could reshape assumptions about how much power capable AI requires. Even if it does not, the fact that Bezos and top venture firms are funding it signals real concern about AI's energy trajectory.

For everyone else, treat this as a science bet to watch rather than a product to plan around. The interesting milestones will be technical: whether Cortex AI demonstrates useful capability at its target power envelope. That is years out, but the direction, doing more with far less energy, is one the whole industry has reason to hope succeeds.

Frequently asked questions

How much did Flourish raise?

Flourish raised $500 million at roughly a $2.5 billion valuation in a round that closed around June 4, 2026. Backers include Jeff Bezos, Lux Capital, and GV.

What is Flourish building?

A system called Cortex AI that uses connectomics, the mapping of real neurons and their connections, to build brain-inspired AI. The goal is capable AI that runs on far less power than today's hardware.

Why is the power target significant?

Cortex AI aims to run in the 20 to 50 watt range, roughly a laptop, versus the kilowatts a server rack draws. That would be an order-of-magnitude efficiency gain and would ease AI's growing energy demands.

Who founded Flourish?

It was co-founded by Thomas Reardon, who created Internet Explorer and later founded CTRL-labs, and Rob Williams, a former Amazon senior executive. Their combined software and neuroscience background helped attract top investors.

#ai#funding

Sources & further reading

Keep reading