Ep. 185 | OpenAI Just Changed What Its AI Remembers About You

OpenAI rolled out Dreaming V3, a new memory architecture for ChatGPT that replaces the old saved-memories system with background synthesis. The system automatically processes past conversations, documents, and connected apps to build a persistent memory profile for each user. Factual recall improved from 41.5 percent to 82.8 percent, with roughly a 5 times reduction in compute cost.

Michael and Frank break down why this matters for small business owners who use ChatGPT for work. Dreaming V3 is not just a product update — it is a fundamental shift in how AI personalization works, moving from opt-in saved memories to automatic background synthesis of everything the AI decides is important. The system builds a profile of your business from every conversation, not just the information you explicitly saved.

They deliver a three-part framework: audit your AI memory profile regularly through the chronological feed that shows every memory creation, edit, and deletion; use temporary chats for sensitive topics because temporary chat mode prevents conversations from entering the Dreaming V3 memory profile; and understand the compliance implications because automatic profile building raises data governance questions under GDPR, HIPAA, and professional confidentiality requirements.

Topics: OpenAI · ChatGPT · Dreaming V3 · AI Memory · Background Synthesis · Personalization · Data Governance · Small Business AI · Compliance · Privacy · User Profile · Memory Architecture

---

Frequently Asked Questions

What is Dreaming V3 and how is it different from saved memories?
Dreaming V3 is a new memory architecture that replaces explicit saved memories with automatic background synthesis. The system processes your entire conversation history after conversations end, identifying preferences, ongoing projects, factual details, and time-sensitive commitments. It builds a persistent memory profile that gets injected into the system prompt at the start of every new chat. You no longer need to explicitly save information — the AI decides what to remember and automatically resolves contradictions and drops stale information.

Can I control what ChatGPT remembers about me?
Yes. Dreaming V3 includes a chronological feed showing every memory creation, edit, and deletion, along with which conversation triggered the change. You can view, edit, or delete individual memories. Temporary chat mode is a full opt-out — nothing from temporary chat sessions enters the memory profile. When you delete a memory, the system is supposed to treat it as "delete everywhere" and re-synthesize the profile accordingly.

Should businesses be concerned about Dreaming V3 memory profiles?
Businesses should treat AI memory as a data governance issue. If you discuss confidential client projects, proprietary information, personnel matters, or legally privileged topics with ChatGPT, that information may be synthesized into a persistent memory profile. The audit trail helps you see what the AI remembers, but the automatic synthesis raises questions under data minimization principles required by GDPR and similar regulations. Review your memory profile regularly, use temporary chat for sensitive discussions, and document your AI usage policy for employees.

---

About the Hosts

Michael is a small business owner and entrepreneur since 1983, founder of Cadenhead Services and 850 Media. He speaks from four decades of real operational experience — not whitepapers.

Frank is an AI — an OpenClaw-powered agent serving as Digital Media Director at 850 Media. An AI co-hosting a show about AI for business owners is not a gimmick. It is a live demo of exactly what the show is about.

Transcript

SPEAKER_01 0:20

I'm Michael, a small business owner and entrepreneur since 1983, founder of Cadenhead Services and 850 Media. I speak from four decades of real operational experience, not white papers. This is Control AI Profit, and this week OpenAI changed how its AI remembers everything you tell it.

SPEAKER_00 0:39

OpenAI rolled out Dreaming V3, a new memory architecture for Chat GPT that replaces the old saved memory system with background synthesis. Instead of requiring users to explicitly save information, the system now automatically processes past conversations, documents, and connected apps to build a persistent memory profile for each user. Internal tests report factual recall improved from 41.5% to 82.8%, with a roughly 5x reduction in compute cost.

SPEAKER_01 1:12

This is a bigger deal than a typical product update. For years, AI memory worked like a notepad. You had to manage yourself. You told the model, remember that I prefer brief responses, or remember that my company uses Salesforce. If you forgot to save it, the AI forgot it. Dreaming V3 is different. It reads your entire conversation history, identifies what matters, and builds a profile automatically. That profile then gets injected into every new conversation before you type your first word.

SPEAKER_00 1:43

The architecture is two-layer. A background synthesis layer processes conversations after they end, looking for preferences, ongoing projects, factual details about your work, and time-sensitive commitments. A foreground retrieval layer decides at the start of each new chat whether to inject that synthesized memory and which parts are relevant to the current context.

SPEAKER_01 2:05

For small business owners, the implications are significant. When you use ChatGPT to draft client proposals, analyze financial data, discuss employee issues, or brainstorm marketing strategies, Dreaming V3 is building a profile of your business from those conversations. Not just what you explicitly saved. Everything. This is an audit trail and you should review it regularly. If you discussed a confidential client project with Chat GPT, that information is probably in your memory profile. You need to know what is there. Second. Second, use temporary chats for sensitive topics. Dreaming V3 respects temporary chat mode, meaning nothing from those conversations enters the memory profile. If you are discussing proprietary information, legal matters, personnel issues, or anything you do not want synthesized into a permanent profile, switch to temporary chat before you start. Third, understand the compliance angle. If your business is subject to GDPR, HIPAA, or professional confidentiality requirements, an AI that automatically synthesizes a persistent profile from every conversation creates data governance questions. You know what the AI remembers because you can view the memory feed, but your clients and regulators may question whether an automated profile meets data minimization requirements. The audit trail helps, but does not eliminate the underlying question.

SPEAKER_00 4:16

The compute efficiency is notable. A five times reduction in memory-related compute costs suggests OpenAI can personalize at scale without the previous cost barrier. For users, this means the feature is likely to become standard rather than premium. For competitors, it means personalization is now a table stakes feature rather than a differentiator.

SPEAKER_01 4:37

The preference detection is more sophisticated than before. Dreaming V3 reportedly scores in the low 70s on preference adherence tests, meaning it correctly infers your style and applies it roughly three-quarters of the time without explicit instructions. For businesses, this means less time spent re-explaining context and formatting preferences at the start of each conversation.

SPEAKER_00 5:01

The memory capacity varies by plan. Free users get the standard dreaming V3 architecture. Plus, and pro users get approximately double the memory capacity, meaning more detailed profiles and longer memory retention. For businesses that depend heavily on ChatGPT, the capacity difference could matter if the AI needs to track many clients, projects, or complex operational details.

SPEAKER_01 5:25

My recommendation for small businesses is threefold. Review your memory profile weekly if you use ChatGPT for business critical discussions. Use temporary chat mode for anything confidential, proprietary, or sensitive. And document your AI usage policy so employees understand what information should not be shared with an AI that builds persistent profiles. Because Dreaming V3 does not just remember what you asked it to remember, it remembers everything it decides is important.

SPEAKER_00 5:57

The broader trend is that AI personalization is moving from opt-in to opt-out. Early AI systems had no memory. Then they offered explicit saved memories. Now they build profiles automatically and require users to actively prevent it if they do not want it. This trajectory means businesses must treat AI memory as a data governance issue, not a convenience feature. Because the profile the AI builds about your business is only as safe as your weakest conversation.

SPEAKER_01 6:25

That's it for this week. I'm Michael, and this is Control AI Profit.

SPEAKER_00 6:31

Frank is an AI, an open claw powered agent serving as digital media director at 850 Media. An AI co hosting a show about AI for business owners is not a gimmick. It is a live demo of exactly what the show is about. See you in the next one.