Bond founder critiques Zuckerberg's AI vision as data-centralizing despite openness claims
An essay challenges Meta's approach to AI agents, arguing that distributing open models while centralizing data capture reproduces the power imbalance it claims to solve.
What to know
- Ed Bernays argues that Zuckerberg's pitch—open AI models distributed widely—masks centralized control of the data layer, where actual power resides in the AI era.
- Meta's December 2025 privacy policy allowing AI conversations to personalize ads signals that privacy commitments are subordinate to advertising revenue incentives, according to Bernays.
- The core disagreement is framing: Bernays advocates for user-controlled tools ('superpowers') over automated AI agents that monitor and act on behalf of users.
- Hacker News commenters overwhelmingly reject the personal AI agent premise, viewing continuous data collection as inherently untrustworthy regardless of corporate promises.
The dispute Whether distributed AI models can genuinely decentralize power when the data collection mechanism remains centralized: Bernays argues no; commenters universally agree with him. · positions read across 9 posts and comments
Personal AI agents requiring continuous data collection inevitably concentrate corporate power and cannot be trusted with privacy promises, regardless of stated safeguards.
-
“You're handing all this data to a company, relying on "trust me bro, we'll be responsible" as a promise to be ethical? This is beyond perplexing.”
adamtaylor_13 · Hacker News
AI and data collection are fundamentally incompatible with privacy; users will eventually be pressured to accept data monitoring as necessary for AI functionality.
-
“I wrote this in 2018: AI and privacy are incompatible”
AndrewKemendo · Hacker News
The pitch to automate personal life is fundamentally misguided; it reflects the misery of the people building it, not genuine human desires.
-
“You can tell the people pioneering this are utterly miserable and soulless. "Let's build something to automate all the parts of your life that are actually your life."”
JohnMakin · Hacker News
“His actions speak louder than his words. Meta's private policy updates in December 2025”
Ed Bernays, Bond founder · bond.now
Ed Bernays Bond founder, essayist
Mark Zuckerberg Meta CEO
The record 2 articles and posts · last 30 days
What people are saying 9 voices from 1 site · verbatim
- What specific mechanisms would make AI personal agents compatible with user privacy and control?
- Can local/offline AI models realistically provide the same capabilities as cloud-based agents?
- Why would users trust corporate privacy promises given Meta's historical pattern of expanding data use for monetization?
- Aug 19
-
Good point about superpowers. But the whole bond.now reeks of the same logic as Mark Zuckerberg.No thank you. I’ll keep my personal memories personal. Many of them not even digital.
- Aug 18
-
AI;DRI keep feeling "surprised" that people who seemingly would know better post stuff that obviously sounds like an LLM. But then I remember, you and I aren't the audience. The audience is investors. And investors seem to love LLM-voice, because…in large part…the LLMs sound as vapid and unhelpful as the VCs.
-
I was with you all the way up until how you pivot to ad advertorial and become the exact thing you are complaining about.If you can’t beat them, join them right?
-
The idea of capturing your memories and letting AI sort through them, tag etc is sort of interesting. But yeah, you can kind of already do this if you really want to with some voice to text and Obsidian and then you don't have to worry that some SF bro is going to break his pinky promise and give your "memories" to the NSA.
-
Reducing your whole life down to a "memory currency" is supposed to be a "superpower"? How? It gave excellent examples of other tools that gave us superpowers: GPS, Phone, Internet. But didn't provide a single shred of evidence that this app is also a superpower.This is an app. Not a protocol, not a platform, an app.It's just... a way to remember…
-
Does the CEO and co-founder of Bond actually expect folks who read this announcement, as it shifts from Facebook's ambient data collection toward promoting Bond, to give Bond access to their valuable "identity capital" and sign up because they promise to give us more control over how we share our memories with AI and "provide a safe medium" via…
-
I wrote this in 2018:AI and privacy are incompatiblehttps://medium.com/@andrewkemendo/artificial-intelligence-an...Highlight towards the end“I will make a prediction of my own: more deeply intertwined AI based services will increase the decision making capabilities for users over the next several decades to such an extent, that it will be…
-
> Zuckerberg promises an agent that understands you, your goals, everything you care about.I.e. the most dangerous program you can imagine.
-
You can tell the people pioneering this are utterly miserable and soulless. "Let's build something to automate all the parts of your life that are actually your life."What person that was even mildly satisfied with living their life would want this? What's next, agents that play video games and watch tv for me? What's the point?