chwahoo

> The problem is that these instructions are not applied after the model has finished doing the work, it becomes part of the same work

Would love to see some data backing how strong the effect is.

Hah, not meta or anything related

Perhaps I'm ignorant.

I know some of the criticism of Meta: many people don't like the way their products are optimized for engagement. I've heard about their weird AI bots interacting on their platform as if they were people. And I know people of all political stripes have had complaints about content moderation and their algorithm.

But all of that is within the bounds of the law and their terms of service.

None of it would remotely approach something like: bypassing the well-advertised features in the glasses that show when the camera is in use and secretly recording things to train AI. It's hard to imagine any company's lawyers approving something like that. (this sounds like what many commenters believe is happening)

FWIW, I suspect this is the relevant section of the Privacy policy:

> "When you use the Meta AI service on your AI Glasses (if available for your device), we use your information, like Media and audio recordings of your voice to provide the service."

from: https://www.meta.com/legal/privacy-policy/

if so, "to provide the service" is doing a lot of work

They're sunglasses so I mainly wear them outside.

I think the most likely case is: this company is labeling images from meta AI use from people who opted-in to share their data with Meta.

It's certainly possible that it's something much more surprising / sinister, but there is a fairly logical combination of settings that I could see a company could argue lets them use the data for training.

I'm also very certain that few users with these settings would expect the images to be shown to actual people, so I'm not defending Meta.

I'll confess that I like my Meta Ray Ban glasses: I love using them to listen to podcasts at the pool/beach, while riding my bike, and it's cool to snap a quick picture of my kids without pulling out my phone.

I wish this article (or Meta) were a bit clearer about the specific connection between the device settings and use and when humans get access to the images.

My settings are:

- [OFF] "Share additional data" - Share data about your Meta devices to help improve Meta products.

- [OFF] "Cloud media" - Allow your photos and videos to be sent to Meta's cloud for processing and temporary storage.

I'm not sure whether my settings would prevent my media from being used as described in the article.

Also, it's not clear which data is being used for training:

- random photos / videos taken

- only use of "Meta AI" (e.g., "Hey Meta, can you translate this sign")

As much as I've liked my Meta Ray Ban's I'm going to need clarity here before I continue using them.

TBH, if it were only use of Meta AI, I'd "get it" but probably turn that feature off (I barely use it as-is).

I haven't closely followed distributed protocols like this in years so might just be just behind the times, but I can't recall ever seeing "Pro Censorship" as a badge on a technology.

Seems to fit with real-world trends that confuse and frustrate me. I'm not saying we need a censorship-proof world, just that I find this quite jarring. I suspect I might be missing some context.

Russian hacking is very real. Highly recommend https://www.amazon.com/Sandworm-Cyberwar-Kremlins-Dangerous-...

I love playing with my mirrorless camera and lenses, but I'm becoming more and more convinced that it's a risky proposition "investing" in a bunch of expensive camera gear (which traditionally holds it's value better than most gadgets) when computational methods will soon evaporate the advantages of bigger sensors / faster glass.

"Furthermore", "novel", "distinct" are scientific jargon that reduce readability? Seems like a non-problem to me.

Another link of interest: http://www.pl-enthusiast.net/2017/05/24/jean-sammet-a-rememb...

You might look at youtube red, which is included with the google music streaming service. Will let you download videos to your mobile device for offline viewing. Combined with your youtube "to watch" list, it works pretty well. Unfortunately doesn't support an "add on desktop -> auto download mobile" workflow.

Can people suggest companies that better balance the factors that I've seen mentioned in this thread:

- Provide good pay/benefits

- Provide a positive culture for employees

- Employees are encouraged to have a life outside work

- Are run in such a way that good decisions are made and the company seems headed for continued success

- Look good on your resume

You're misreading. Everyone they hire is intended to be a leader, and being right is one of the qualities they care about in employees. Not "management is usually right, so trust them".

if the company tanks right after you exercise, can't you deduct the losses?

Kitsune provides useful abstractions for modifying your program for runtime updating and tools for automating state transformation between versions of your program.

It would be entirely reasonable to implement DSU all within an app's own codebase, particularly until there's a production-ready library/tool-set. The downside would be that you'd probably end up re-implementing a lot of what Kitsune provides.

For our purposes (evaluating Kitsune-style updating on a variety of server programs), it made sense that we'd want to have a common toolset that we applied to all of the programs.

This was work done as part of my phd thesis. You can take the code and samples that are here https://github.com/kitsune-dsu and play with them. However, I wouldn't consider what has been released ready for production use. The papers were the main product of this research and are your best resource if you're interested: http://www.cs.umd.edu/~hayden/papers/kitsune-draft.pdf

I moved on (graduated!) from the project in 2012 and the code that has been released is pretty much where I left it at that time. An undergrad collaborator was doing neat work on updating Tor, so that code continued to evolve a bit after I left.

I think there may still be folks at UMD working in some ways with Kitsune, but I'm not up on the details.

The paper describes the main changes that you have to make to make your program updatable and is the best resource.

You can find all the modified programs (redis, etc) here: https://github.com/kitsune-dsu

But as she points out, FizzBuzz doesn't do a good job in telling you if you've found one of these developers because the context of interviewing is different than the context of banging out code on the job.

It may be that connections though xfinitywifi don't count against the neighbors allocated bandwidth and won't degrade their service since their router and connection support more than their allocated bandwidth. Can anyone confirm if this is the case?

Their solution stops the reloads for each browser and has the benefit of alerting the user that something's wrong, without hammering "somewhere".

Who's the steward of your language of choice?

I like reading a bit of prose about when a tool doesn't fit someone's use case. Nim has a pretty interesting approach to data sharing in concurrency, and it's more interesting to read a story of someone who got bit by it than to read a table of language features and wonder when it might prove to be a problem.

Out of curiosity, what tool did you use to sketch the various architectures for your post?

The vast, vast, vast majority of people who are born with every advantage don't accomplish anything worthy of a profile like this. And I don't dispute that, if she'd been born in a starving village in a third-world country (or probably even to poor parents in the US), its very unlikely that that she would have done these particular things.

It seems that the only criteria for interestingness or laudability that you'll accept is to have overcome underprivileged circumstances. I think that's a perfectly fine factor, but I'm personally interested in reading profiles of people for LOTS of other reasons: impressive accomplishments, impressive problem solving, overcoming/motivated by bad things happening to you or your close family, personal struggle, unusual ideas / perspective, hard work, humor, ... and many more (some that I'm not even aware of).

If a person's story has the right weighting of factors like that, I'm interested. And Martine Rothblatt's story was WELL OVER the bar for being someone who I'd want to learn about.

Interesting, I've heard people talk about transhumanism and singularitarianism as similar to religions/cults, but those philosophies aren't usually presented with an explicitly spiritual/religious component. It sounds from the article like the Terasim Movement does try to add a religious component (the profile suggests elements of Judaism). A light skim over the Terasim Movement website (http://www.terasemmovementfoundation.com) doesn't seem to reveal much about that religious component, although this video (http://vimeo.com/100518959) has a nebulously "spiritual" feel to it.

Do you think this cult is dangerous because you believe all cults/religions are, or do you see something particularly worrisome here?

Wow, where did that come from?!? Was starting her workday at 2am everyday part of her "luck"? From this profile, it sounds like she's worked very hard, she's smart, she's sacrificed living on very little as a grad student, she's taken risks, and she's been driven to do work that's important too her (developing technology to save her daughter).

Sure, tons of people live fortunate lives by skating forward on the momentum of a privileged upbringing. But I get exactly the opposite vibe from this article.

I'm surprised you were downvoted. I think this suggestion makes a ton of sense. It's certainly the shortest path to having a significantly improved first-party language for Android and Google does have a working relationship with Jetbrains.

I think it's very unlikely that this is in the works since, if it were, you'd expect Google to have thrown significant weight towards kotlin development (https://github.com/JetBrains/kotlin/graphs/contributors).

I'm not seeing the javascript-alike-ness. What caused that connection to jump out at you?

I see the standard static FP features (from ML, Haskell, Scala, F#) with the syntactic flavor or Rust and some C# tossed in.

I don't think his point is as broad as you're claiming. He's saying that golang application code is complex due to (for example) lack of generics because the language lacks that complexity.