HN

Comments (151)

• Menu

  piazz
  I’m pretty frustrated with Muse and the last thing I want to be doing with my free time is defending Meta, but this is such clickbait.Point by point:> “OMG you can jailbreak it and get it to spill its VM”This is the whole point; any content on the VM is yours. It runs in an isolated sandboxed VM separate from stored credentials etc; this is effectively your own computer. You don’t have to trick it.> It collects dossiers on your contactsThese are more text files that live on your private VM, alongside memory.md, etc. Do you want your secretary to forget every person you contact every day?> It accessed Messages without full disk accessThis whole story never made sense or was substantiated. Full disk access is an OS level security boundary; the user had to switch this on.> It sold some guys stuff for too cheap and gave out his addressOK this one I basically believe, haha. Because this is the problem with Muse: the LLM is just too dumb to perform complex tasks effectively in many cases.
• Menu

  jagermo
  I get the appeal of agents, I do. This is the future stuff we always wanted. But I cannot get myself to give one of these things access to my bank account or allow it to do price comparsion and shopping without oversight. Or access to my email or chat history.I just do not trust any of them, not with my money or with access to my conversations.
• Menu

  chadd
  It's very simple. There is ZERO chance the worlds biggest advertising companies will refrain, long-term, from using your most intimate secrets, gathered through your many conversations with their AI personal assistants, to sell you things.
• jjcm
  I don't think anyone is arguing it isn't one. The fact that all of these agents have provisions for keeping .env vars and credit cards out of their chain of thought and responses is evidence enough of that - they know full well the info they're handing can easily be leaked. Notably though PII is absent from that, since without that information the agent just isn't valuable.What most people are arguing is whether or not it's worth it. These things are essentially executive assistants, which historically have also been massive vectors for security leaks. EAs know the most about you, your behaviors, and your motivations. Still an EA is almost always worth it if you can afford one. Given now the cost is free, Muse is something I've both used and recommended. It's worth the risk for me.at the cost of all your data and probably your eternal soul
• Menu

  alistairSH
  Are all of these just flavors of "the agent has the same permissions as the user"? Not that I want to trust Meta with anything, but I'm guessing it asked for "root" access and the users granted it...?
• ygjb
  I continue to believe that Meta is the company that is building products I want from a company I have absolutely zero desire to do business with. I put Google in the same category - I used to be a google fanboy, but that subsided over the years as Google grew and killed products I used and liked. Now I only use their products because of inertia and network effect.I don't trust any of the hyperscalers to place me as a user first, I expect them to guide me into a stall, strap on a feedbag, and start the value extraction process while they build the meatrix[1] around me.I want tools like Muse, and Meta Glasses, Google Gemini and a horizontally integrated AI agent, but I want it on my terms, where I am in control and accountable, and without a layer of data collection and harvesting on top of it.Fortunately I am skilled enough to assemble most of what I want, but I fear a world where people who lack the skills and resources I have available have no choice but to onboard and give up control in order to remain competitive, employed, and connected in the world to come :([1] https://www.youtube.com/watch?v=rEkc70ztOrc
• geophile
  A tangent, but todays’s NYT Daily podcast, about one journalist’s experience with Muse, was pretty irresponsible in not covering these problems.
• Menu

  matltc
  Couldn't pay me to use this junk.Trying to think of a number that I would consider it, and honestly $1k/month wouldn't convince me unless- I have root on device- device is on its own vlan, fully segmented- !(SIM || 5G antenna)- no google/apple id authenticated on device- terminate agreement at any time without costI'm probably forgetting/not aware of ten things I should consider.
• autoexec
  Facebook has never been shy about what they are and how little they care about you. They've demonstrated it over and over and over. At this point if you use their products you're asking for what you get. They should still be sued into the ground for lying to people about the privacy of Muse (and everything else), but why the hell is anyone still using their products at all? How has this company got so many people acting like battered spouses? Facebook is never going to stop hurting you. It's time to leave Facebook for good.
• Menu

  DebtDeflation
  >When one tech YouTuber put Muse in charge of their Facebook Marketplace sales, it sold his stuff way below acceptable ratesWhy are we depending on LLM "reasoning" to negotiate a price rather than just having the user enter a lowest acceptable price deterministically?
• Menu

  arshxyz
  > Meta’s new general-purpose AI agent Muse sent him an unsolicited notification referencing a thread between him and a co-worker over Apple Messages. Aten said he never granted Muse permissions to read his messagesCan someone explain how this is possible? If an app can do this without Full Disk Access or a popup of some kind that's a way bigger lapse on Apple's part than Meta's.
• sdcfgy
  Isn't that Meta's entire product line?
• Menu

  whycome
  It’s interesting that the only ads I’ve seen for Muse don’t mention meta at all.
• Menu

  honestduane
  But isn't Muse just a fork of OpenClaw?
• Menu

  labrador
  I can finally relax and let a random number generator make my decisions for me
• chasing
  Have we learned nothing?The MOMENT they can, Meta will take all data they have about you and sell it to anyone who wants it, including to people who wish to do you harm.Honestly, this is where actual government regulations with teeth would make me significantly more comfortable. But self-regulated? Absolutely not.
• compiler-guy
  Just remember that Meta (née Facebook) changed the entire Facebook feature set and feed style in search of greater engagement and clickbait so that you would do what was best for it rather than satisfy your preferences. And if you liked it more the old way, tough luck. Your feed is now a stream of rage-bait, half-dressed women, and engagement slop instead of a way to keep in touch with acquaintances.No way I would allow them to develop the agent that runs my life. Even if it works well now, the rug pull is absolutely inevitable.[I was never all that into facebook, but it was nice to get an update on an old high-school friends once every month or two. Now its completely unviewable.]
• fridder
  This shouldn't be a surprise to anyone. Why would you trust Meta with privacy and security?
• Menu

  jeanpah
  Again? This seems very intentional at this point
• coliveira
  And it's all by design, Meta and its founder have a long history of releasing products with security "flaws" that are immediately used by them to collect vast amounts of information about their victims.
• chrisjj
  > Muse consistently creates detailed profiles of all your friends, family, colleagues, “collaborators,” and people you “follow.” Obviously much of that information is necessary for the agent to get to “know” you, but this being Meta, people are understandably uncomfortable with this sort of massive ramp up of data collectionPeople who give their data to Meta are uncomfortable with Meta collecting that data.Hmm.
• cdrnsf
  Everything Meta makes is a privacy and security dumpster fire.
• Menu

  Havoc
  The model is decent and cheap under contributor version but no way I’m letting meta AI anywhere near anything that matters in my life.> Zuck: They "trust me"> Zuck: Dumb fucks.
• measurablefunc
  Every social media & AI company is also a surveillance company. Targeted advertising doesn't work w/o mountains of behavioral data aggregated across all of Meta's & Google's software "products".
• Menu

  Razengan
  Does anyone remotely interested in AI use Muse out of explicit choice?Facebook is like Microsoft at this point: The thing your grandparents use.Even if they make something technically superior for a while, like what the Zune was to the iPad, tHey'll never really be cool.and they'll certainly never be trusted.
• Menu

  otikik
  Oh Meta not giving a damn about privacy and security? Say it ain't so.
• anon
  undefined
• Menu

  hannofcart
  I understand that a typical HN user is very different from the average person.However, surely by now even the lay people who have seen the testimonies before Congress, the lawsuits and the excesses by Meta execs over and over again ought to be atleast somewhat wary of handing them more personal data?I think the average person has very similar self preservation instincts as the rest of us. So it can't be that. Which leaves the fact that there are lot of people who don't know about the above mentioned scandals galore that Meta has been involved in?
• anon
  undefined
• ChrisArchitect
  Since this is mostly a collection of links to previously discussed articles:Related:Updates to Full Disk Access in macOShttps://news.ycombinator.com/item?id=49937631Meta’s Muse has a serious 0-dayhttps://news.ycombinator.com/item?id=49802030Unsurprisingly, Meta's new Muse AI agent blatantly ignores users permissionshttps://news.ycombinator.com/item?id=49893709Maybe don't let Muse run your Facebook Marketplace accounthttps://news.ycombinator.com/item?id=49875006What Meta got right with Musehttps://news.ycombinator.com/item?id=49946526Muse – Meta’s personal AI agenthttps://news.ycombinator.com/item?id=49615537
• Menu

  nekusar
  "People just submitted it. I don't know why. They 'trust me'. Dumb fucks."-Mark Zuckerberg
• magus-stoopr
  [dead]
• ContinuityLab
  [dead]