Weren’t some people, unironically, expecting AgI announcement for GPT-5. Like I have heard a water cooler (well, coffee machine) conversation about how OpenAI master plan is to release GPT-5 and invoke the AGI clause in their contract with Microsoft. I was shaking my head so hard,
They are both using the "capitalist" definition of AGI, that is "an AI system that can generate at least $100 billion in profits". I think it's short for "A Gazillion Idiots"...
It is actually incredible how they managed to find an even
more unscientific definition than "can perform a majority of economically useful tasks." At least that definition requires a little thought to recognize it has problems[1]. $100bn in profits is just cartoonishly dumb, like you asked a high schooler to come up with a definition.
[1] If a computer can perform the task its economic usefulness drops to near zero, and new economically useful tasks which computers can't do will take its place.
If there's actually any proprietary rockety data, maybe. Without knowing what data went into the fine-tune there's no way to tell. This could be a "internal procedures chatbot" or an "onboarding chatbot" where new people can ask where the coolest watercooler in the company is.
In my experience post-training mainly deals with "how" the model displays whatever data ("knowledge") it spits out. Having it learn new data (say the number of screws on the new supersecretengine_v4_final_FINAL (1).pdf) is often time hit and miss.
You'd get much better results with having some sort of RAG / MCP (tools) integration do the actual digging, and the model just synthesising / summarising the results.
Or, since we're apparently playing the game of maybes in this thread, maybe the LLM was only trained on the teams grandmothers' spaghetti recipes, so that new hires can learn to make the best bolognese sauce.
I didn't miss anything in the wordplay*, it was obvious. (As are the initials, an extra pun).
I put quotemarks around "flamethrower" because that's what it was originally sold as before obvious and predictable legal issues with real flamethrowers and the fact it was obviously mimicing the prop in Spaceballs.
My point is: neither weed burners nor actual flamethrowers have anything to do with digging tunnels nor any adjacent aspect of civil engineering.
Boring as the noun, not adjective. Also, Tesla was named that before Musk was involved, so it’s not his humor involved in naming both. Nikola Tesla is known for a lot more than just Tesla coils.
ITAR (International Trafficking in Arms Regulation) is paranoid. Every single specific person that knows even dual-use information, such as composite wing design, must be individually authorized. I’ve been asked to leave the room when my girlfriend, who works for a passenger aircraft manufacturer, was designing a repair for a plane I have literally flew on.
It doesn’t matter how the person got access to dual-use info, like basically everything to do with large rockets, it’s 100% forbidden.
That's the people aspect of it, but what about the technical aspect of it? Can I store ITAR restricted information in plaintext on a thumb drive if I think it's safe?
> It's a speaker system. It plays sound. Why could it possibly have AI, tracking, or ad delivery?
To recognize what you listen to, build a profile, feed it back to Samsung, which will use it in deciding what crap to display on your Samsung TV (and any other devices) associated to the same profile. For all we know it's even listening to your conversation in the room, I mean, it's Samsung - they literally do this:
How much benefit could that bring versus burning reputation and losing it all? These companies are so big and powerful but time and time again they keep on forgetting that they can't exist without the users and when users start leaving it's hard to reverse that trend.
It's so out in the open if you know, or more likely, worked in media advertising.
Their competitor, Vizio, owns iSpot[1] which is, in my opinion, the best in the space.
Samba TV[2] is it's nearest competitor and they have their hooks into 24 Smart TV brands globally[3]. These brands are listed on their website as Philips, Sony, Toshiba, beko, Magnavox, TCL, Grundig, Sanyo, AOC, Seiki, Element, Sharp, Westinghouse, Vestel, Panasonic, Hitachi, Finlux, Telefunken, Digihome, JVC, Luxor, Techwood, and Regal.
There is no reputation to burn, they're well known to do this kind of stuff by anyone bothering to look it up, and nearly nobody looks it up anyway.
It's a pity because I liked some of their hardware in the past (an NX camera I still have, hard disks back in the IDE stone age, 3 LCD screens back from when they were a novelty - they only had a VGA connector) but I just stay away from them now. But 0.01% of their customers staying away is completely insignificant when they consider the profit opportunity of violating our privacy.
Come on, did you read more than just the headlines?
> Samsung's spokeswoman continued: " Should consumers enable the voice recognition capability, the voice data consists of TV commands, or search sentences, only. Users can easily recognize if the voice recognition feature is activated because a microphone icon appears on the screen."
So it is not like it was listening without your knowledge. Only when you use the voice features is the data being sent over. Like with every other online service. As much as I don't like samsung, this is a bullshit reason to hate them.
And why provide two links basically saying the same about the same story?
Their competitor, Vizio, owns https://www.ispot.tv/ which is used for ad delivery tracking.
It's much more reliable and precise than the familiar Nielsen ratings: since you know the total audience of X% TV households in a zipcode (which you know demographics of race/income/household size based upon), and Vizio TVs account for Y% of all TVs sold for households with incomes between A and B, and C and D you can get a confidence interval of how many people ACTUALLY saw your TV advertisement.
Samsung was/is probably trying to do something similar: All sound in your TV pipes through their home theater system, so they can "Shazam" whatever media you're watching, regardless of the source (OTT, OTA, hell even YouTube or a Downloaded Torrent on your laptop hooked up via HDMI) and phone home.
on android you can install SoniControl Firewall to "see" the ultrasonics in your house. Try it with all tvs and things off, then try it with the TV on, youtube videos, and so on.
Pixel tracking works better if the TV is connected to the internet. I remember samsung as one of the companies, where, if your TV was not ever given a wifi connection, it would attempt to connect to any open network to do what it needed to do. This sounds unlawful, so i don't know the veracity, but anyhow - if the TV is online, it can just send a half dozen pixels at known locations back home and there is a database of "content pixels at timestamps" and they match the half dozen pixel values to the database and know what you're watching to some degree of certitude.
but for things like dumb panels older TVs and the like, ultrasonics still work.
You can just use regular math to do this. We've been doing it for 30 years now. You don't need a trumped up overpriced garbage LLM to do anything for you here.
Didn't know that, thanks. Then speakers are actually a pretty big data source. I bet most people don't assume their speakers can be listening. I wonder if you can get internet connection over bluetooth aux or what'd be the best way to get someone to let you send data home on a speaker.
i did some cursory digging, but i don't really want to read the A2DP or AVRCP specifications to see how much data is allowed in the non-audio payload. Besides, PAN exists, but i imagine you have to do something on your phone to allow it.
Most of these expensive things also have wifi, though, don't they?
> Connect your devices and control everything with our soundbar that integrates your favorite voice assistants and smart services like Built-in Alexa², Chromecast³, Airplay 2⁴ and more.
Few things over the past few years have infuriated me as much as tracking and advertising being introduced at the OS level, especially on TVs. I'm looking at you, LG! I will gladly pay more for a TV that doesn't try to advertise Roku's streaming service to me or track my kids' watch history. Seems like they are few and far between, though.
The best thing we have been able to come up with is leaving the TV itself disconnected from the WiFi and using an Apple TV for smart features/streaming. I'm sure they're still gathering data but it's at least not as blatant. It's a real crapfest for the consumer at the moment.
This is sound advice for keeping yourself free from malware as well. Many of these TVs end up running super vulnerable junk that doesn’t get updated and has known exploits.
I’ve had two devices end up with malware like this. A Sony blue ray player that was uploading 2gig a month before I caught it and a Samsung tv.
It’s worth mentioning you have to block or change WiFi credentials. The device with malware may attempt to connect to any known wifi even if you disable it on the device. I get 45000 auth attempts a day from my tv.
> I will gladly pay more for a TV that doesn't try to advertise Roku's streaming service to me or track my kids' watch history. Seems like they are few and far between, though.
This just swaps one locked-down company for another. You're still at the mercy of a giant corp, and worse it's unlikely to work well with my linux laptop and Android phone whereas at least Samsung tries (and often fails). A better solution is needed. I buy Sceptre TVs when I can, though for a "big screen" there aren't great options.
Yeah, we do use Apple TV because at the very least if they are collecting our data, they're not using it to advertise directly to us on the same device. My parents have a Roku TV and the number of ads it serves up directly on the device leave me feeling nauseous.
