OpenAI CEO’s Ties to Chip Startup Ensnared by National Security Review

OpenAI's Chip Deal Stalled by National Security Review: Ambitious Project on Hold
May 21, 2024

‍

‍

A 2019 agreement for AI research giant OpenAI to purchase $51 million worth of neuromorphic AI chips from Silicon Valley startup Rain Neuromorphics has come under scrutiny - highlighting the interplay between OpenAI leadership’s personal investments and national security priorities amidst breakneck competition to develop specialized hardware for artificial intelligence workloads.

The deal, first unveiled last spring, intersected a maelstrom of competing storylines centering on OpenAI as it races to cement status as a leader in next-generation AI capabilities.

From turbulence in its executive leadership, to fights over its direction and reliance on cloud infrastructure, to eyebrow-raising competition moves - the saga encapsulates many of the pressures bearing down on the high-profile nonprofit as attention reaches fever pitch following its release of ChatGPT at the end of last year.

Central to it all has sat CEO Sam Altman all along, whose entanglements with Rain Neuromorphics probe more profound questions about OpenAI’s plan to satisfy skyrocketing compute demands in the years ahead.

Altman’s Personal Ties to Rain Date Back Years Before OpenAI Deal

While only surfacing more recently, Altman’s ties to Rain in fact originate from shortly after the fabled Y Combinator boss first helped incorporate OpenAI as a nonprofit back in 2015.

Just a year later in 2016, Altman would join a seed funding round that funneled $1.1M into Rain Neuromorphics - taking a personal stake in the nascent startup working to replicate features of human cognition via a novel AI chip architecture.

Hailing originally from the University of Utah but operating today close by OpenAI headquarters in San Francisco, Rain Neuromorphics has spent close to a decade developing neuromorphic processing units designed to deliver radically improved efficiency for artificial intelligence workloads.

Emulating the parallel distributed processing of the human brain, the startup aims to push a new breed of proprietary AI chips specialized for inference rather than training, targeting applications in autonomous mobility and edge intelligence use cases running on extremely restricted power budgets.

While still likely years away from commercialization, leaked specs from Rain suggest its NPU architecture could theoretically attain computation performance up to 100x greater than leading AI accelerators today, while requiring over 10,000x less power thanks to bio-inspired efficiencies.

If claims hold up, that combination of speed and efficiency could enable a breakthrough class of perpetually inference-ready AI to be embedded cost-effectively virtually everywhere - from smartphones and laptops, to smart home devices, drones, autonomous cars, robots and industrial IoT infrastructure.

But between the technical challenges of commercializing unproven architectures and now the glare of a national security spotlight, Rain still faces a rocky path to actualizing any of those use cases at scale over the next few years.

OpenAI Deal Preceded Shift to Reliance on Microsoft Cloud Infrastructure

Of course, Altman wouldn’t be alone in seeing long-run potential upside beyond Rain’s immediate commercial viability challenges.

OpenAI themselves clearly bought into some element of Rain’s vision back in 2019 - signing an agreement in May of that year to purchase Rain Neuromorphic’s first $51 million worth of chips off the factory line for integration into OpenAI’s workflows.

Notably however, the 2019 deal came months before OpenAI entered into an exclusive partnership with Microsoft - announced that same summer - which saw OpenAI migrate the bulk of its AI workloads over to Azure’s cloud platform and infrastructure services.

OpenAI had relied heavily on GPUs provided by Nvidia and Google’s Tensor Processing Units prior, but collaboration with Microsoft gave them access to cutting edge supercomputing capabilities leveraging tens of thousands of proprietary Azure machine learning accelerators.

So while representing a vote of confidence in Rain’s technology, OpenAI’s commitments unsurprisingly appear to have pivoted substantially towards reliance on Microsoft’s more mature, battle-hardened AI hardware capabilities in the years since - likely diminishing any short term need for Rain’s exploratory neuromorphic offerings.

Still, with Sam Altman implicated financially and professionally all along, his continued proximity to Rain always threatened to reinsert complications even amidst OpenAI’s quickly escalating ambitions today.

Sure enough, a string of events late last year would sharply thrust Rain Neuromorphics back into uncomfortable limelight.

Government National Security Review Forces Saudi Investor Out

The impetus stems from national security concerns triggered over one specific Rain Neuromorphic investor: Saudi Arabian sovereign wealth fund and venture capital outfit Prosperity7 Ventures.

As part of a 2020 funding round, Prosperity7 secured an equity position in Rain, instantly flagging regulatory alarm bells for the Committee on Foreign Investment in the United States (CFIUS) - a governmental organization tasked with staving off security threats from foreign involvement in US companies.

CFIUS ultimately took the rare step of requiring mandatory divestment from Rain by Prosperity7, compelled by anxieties that Saudi proximity to bleeding edge semiconductor R&D posed too great a potential conduit for cyber espionage or future technology transfer risks back to the Gulf monarchy.

Requirements to pull Saudi funding from American semiconductor firms have sharply escalated recently, as CFIUS battles to balance economic exchange interests against ever more brazen attempts by adversaries to mine US innovation through third party affiliations worldwide.

And in Rain’s case, jeopardy likely accelerated given profound security sensitivities around brain-inspired AI hardware - representing a rapidly militarizing sub-sector as advanced militaries move aggressively to capitalize on AI supremacy.

The anxious scrutiny has already stymied multiple Chinese semiconductor manufacturers in recent years, and will only heighten around novel AI chips able to hugely magnify the power efficiency - and thus proliferation - of machine learning applications.

In Rain’s case though, former CEO Jim Trounson pushed back firmly against notions any valuable IP ever reached current or former investors, asserting robust controls always protected core data and design documents throughout the company’s history.

Nonetheless, resolving the forced removal of Prosperity7 funding promised to hamper Rain’s development plans regardless by early 2023 - likely diminishing viability of the original 2019 supply pact with OpenAI in the process.

RAIN Leadership Shakeup Adds Further Uncertainty Around OpenAI Vision

And internally, Rain Neuromorphics entered 2023 rudderless following the resignation of Trounson as CEO right as the CFIUS ruling landed late last year.

Citing personal reasons, his exit saw Rain suddenly leaderless just as it readied batches of prototype NPUs for select partners, on the cusp of unveiling working demonstrations headed into 2023.

Rain has since engaged an executive search firm to identify seasoned industry veterans to take up the mantle, but filling the CEO chair may prove challenging given both the technical novelty and now regulatory overhang Rain operates under.

All told, with the startup’s roadmap now in flux, delivering market-ready neuromorphic products at scale looks decidedly further away. And hopes of funneling chips to OpenAI as perhaps Rain’s highest profile early customer fade equally dim - even with Altman’s fingerprints still all over both companies.

Still, Rain’s promise and the recent turbulence each speak to larger forces bearing down on OpenAI - crystallizing hardware bottlenecks compounding as competition intensifies to own the future of AI itself.

OpenAI Hardware Constraints Loom Large Over Sam Altman’s Leadership

In fact, strains emerging in Rain’s strategic relevance to OpenAI hint at simmering tensions within OpenAI’s leadership regarding how best to reconcile its relentless need for specialized AI hardware with corporate governance pressures in recent years.

As co-founder and incumbent CEO, one of Altman’s defining tech philosophies has centered firmly around intelligent vertical integration - maintaining control over foundational tooling rather than outsourcing to external cloud providers or merchant semiconductor players.

Altman has repeatedly voiced consternation with shortages and soaring prices as demand for AI specialized chips massively outstrips supply chain capacity currently built by merchant vendors.

And while OpenAI’s alliance with Microsoft Azure unlocked near term access to tens of thousands of cutting edge AI accelerators, Altman has made little secret of longer run apprehensions around growing so beholden to any single external entity - even partners as powerful and deeply invested technologically as Microsoft.

His views crystallized in a proposal earlier this year to possibly spin up a dedicated hardware startup within OpenAI focused solely on designing proprietary AI training and inference chips to cater to their full stack needs.

The philosophy echoes sentiments that catalyzed OpenAI’s original founding in 2015 - with Altman, Elon Musk and early legendary backers pledging to counterbalance reliance on the handful of Big Tech giants dominating bottlenecks around resources critical to AI innovation.

And Altman went as far as floating the need to potentially raise additional funding specifically to incubate the mooted chip design offshoot - likely drawing on lessons from his experiences scaling startups at Y Combinator over the prior decade.

But the board response amounted to a resounding rejection of the plan - rationalizing that attention is better spent on OpenAI’s core competency in algorithmic breakthroughs, rather than dispersing focus across the dizzying complexities of custom hardware R&D too.

The rebuttal showcases a quintessential tension OpenAI leadership continues wrestling to reconcile: power remains firmly consolidated around Altman as CEO to shape OpenAI’s technological direction and partnerships, yet financial governance levers ultimately rest in the hands of the board and key donors like Microsoft.

It’s a dicey juxtaposition almost guaranteeing further strategic conflict even amidst OpenAI’s

MORE FROM JUST THINK AI

MatX: Google Alumni's AI Chip Startup Raises $80M Series A at $300M Valuation

November 23, 2024
MatX: Google Alumni's AI Chip Startup Raises $80M Series A at $300M Valuation
MORE FROM JUST THINK AI

OpenAI's Evidence Deletion: A Bombshell in the AI World

November 20, 2024
OpenAI's Evidence Deletion: A Bombshell in the AI World
MORE FROM JUST THINK AI

OpenAI's Turbulent Beginnings: A Power Struggle That Shaped AI

November 17, 2024
OpenAI's Turbulent Beginnings: A Power Struggle That Shaped AI
Join our newsletter
We will keep you up to date on all the new AI news. No spam we promise
We care about your data in our privacy policy.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Popular
Home
Just Think Chat
Text to Speech
Just Think Art
About
Tools
Teacher Tools
Templates
Resources
Blog
Pricing
Just Think App
Referral Program
Privacy Policy
Terms of Service
Report an Issue
Use Cases
AI Article Writer
AI Content Improver
AI Image Generator
Google Ads Generator
Facebook Ads Generator
Link in Bio
Browse All Tools
Support
Request a Demo
Contact Us
Report a Bug
Report an Outage
Request a new Feature
© 2023 Just Think AI
TermsPrivacy