Good morning, AI enthusiasts. AI chipmaker Cerebras Systems has filed for a $26.6 billion IPO — and OpenAI is woven into nearly every layer of the deal, from being its largest customer to having Sam Altman and other executives as personal angel investors.
The offering drew $10 billion in orders for just $3.5 billion in shares, signaling that investor appetite for AI infrastructure is nowhere near cooling off. If the IPO performs, it could push other AI-adjacent companies to start thinking seriously about their own paths to the public markets.
In today's AI recap:
"The Cerebras IPO is one of those stories that reveals just how tightly interconnected the AI ecosystem has become — the company building the chips and the company running the models are deeply entangled. For professionals watching where AI infrastructure dollars are flowing, this one is worth paying close attention to. — Larry Bruce, BDCbox"
The Recap: AI chipmaker Cerebras Systems has filed to go public, planning to sell 28 million shares at $115–$125 each to raise $3.5 billion at a $26.6 billion valuation — which would make it the largest tech IPO of 2026 so far.
Unpacked:
Bottom line: The Cerebras IPO signals that AI infrastructure is maturing into its own high-stakes investment category, not just a supporting layer for software. If this offering performs, it could pressure other AI-adjacent companies — and eventually the big names like OpenAI and Anthropic — to consider their own paths to the public markets.
"The simultaneous launch of competing enterprise ventures by the two leading AI labs signals that the race for corporate AI adoption has entered a more aggressive phase. For professionals evaluating AI platforms, this shift means vendors are no longer just selling software — they're selling deployment partnerships."
— Larry Bruce, Editor, BDCbox
The Recap: On the same day, OpenAI and Anthropic each unveiled new joint ventures designed to accelerate how large organizations deploy AI — and both are betting on the same hands-on deployment playbook to win enterprise customers.
Unpacked:
Bottom line: For enterprise buyers, this simultaneous push means more hands-on support options — but also more pressure to choose a partner as these ecosystems take shape. The AI platform race is no longer just about models; it's about who shows up to do the work.
The ChatGPT Education Study That Just Got Retracted
From Larry Bruce: "This retraction is a sharp reminder that viral AI research does not always hold up to scrutiny and the damage to the conversation can outlast the correction. For professionals making real decisions based on AI research, knowing how to evaluate the quality of a study matters just as much as reading it." - Larry Bruce, BDCbox
The Recap: A Springer Nature meta-analysis claiming ChatGPT produces a large positive impact on student learning has been retracted nearly a year after publication, after journal editors found discrepancies that undermine its core conclusions.
Unpacked:
Bottom line: The citation trail a flawed study leaves behind does not disappear when the paper does. Applying closer scrutiny to AI research, especially in high-stakes fields like education, is a skill that professionals cannot afford to overlook.
From Larry Bruce:
"The data makes it clear — image AI has become the primary engine pulling users into mobile apps, and that shift is happening faster than most product teams anticipated. For developers and entrepreneurs watching the AI space, understanding where downloads convert into revenue is the real story here."
— Larry Bruce, Editor, BDCbox
The Recap: New data from app intelligence firm Appfigures shows that image AI model launches now drive 6.5x more downloads than traditional chatbot upgrades — a meaningful signal that visual AI features, not text chat, are what get people to actually install apps.
Unpacked:
Bottom line: Image AI features have become the most powerful tool for attracting new users, but winning downloads and winning revenue are two very different challenges. Product teams that close that gap — the way ChatGPT has — will hold a serious edge in the mobile AI market.
DoorDash added a new suite of AI-powered tools for restaurant merchants, including an auto-onboarding feature that scrapes a restaurant's existing website to auto-build its app listing, two photo-editing tools — AI Retouch and AI Replate — that can sharpen, relight, and professionally replate dish photos without altering the food itself, and an automated marketing campaign builder, with merchants in testing seeing order conversion rates of nearly 10% on AI-generated websites.
Researchers revealed that large language models can reconstruct sensitive personal profiles — including political preferences, employment status, age, and financial situation — just from analyzing the stream of Facebook ads a user passively sees, using data from over 435,000 ads collected through the Australian Ad Observatory, with the process being 200x cheaper and 50x faster than human analysis.
Claude powered an unofficial "Notepad++ for Mac" port built by developer Andrey Letov using Anthropic's Claude CLI — but the project sparked a trademark dispute with original Notepad++ creator Don Ho, who called it "misleading, inappropriate, and disrespectful," with the app now being rebranded as "NextPad++" amid ongoing questions about AI-vibe-coded apps riding on established brand names.
NVIDIA uses cuOpt-based agentic workflows with multi-agent LangChain Deep agent orchestration and GPU-accelerated solvers internally to optimize its own supply chains — and because it's open source, the company is making it available to developers with supply chain decisions generated in minutes instead of weeks, along with free developer credits via a preconfigured Brev Launchable environment.
