@TranquilTurbulence@lemmy.zip
Pro tip: If you find large mysterious eggs on a derelict space ship, ALWAYS inspect them very closely. They tend to contain some really cool stuff. Mind-blowing, transcendental best stuff ever. I’m talking, like, way too legit to be legit, feel me? If you see movement inside, be sure to place your head close to the top of the egg. You should try to smell the egg or even give it a lick.
member since 19 May 2024 15:17
comments
- on Shopify CEO says employees' 'slop grenades' are making more work for everyone else in c/technology@beehaw.org · 2d ago
-
on Is AI the ultimate privacy proxy - and the ultimate privacy trap at the same time? in c/privacy@lemmy.ml · 5d ago
I would love to have more of these kinds of conversations, but judging by the downvotes, most Lemmy users disagree.
Anyway, LLMs have changed many things, but source criticism still hasn’t gone anywhere. People use them as if they’re truth engines, ignoring how often they hallucinate. If you’re not qualified to verify the factuality of the answer, the LLM will fool you disturbingly often.
Programming and other technical matters are a bit different though. You don’t have to be an expert of any kind to verify the answer. If the command doesn’t run, the library doesn’t exist or the compiler spits out errors, you’ve just got plenty of verification. The LLM hallucinated some nonsense, and was caught before any real damage could happen.
In other matters, verifying the answer would take time, effort and expertise. Avoiding all of that is kinda the whole point why people outsource their thinking to an LLM. People make these kinds of decisions, because it’s just so convenient. As you’re avoiding modern internet cancer by letting the crawlers take the hit on your behalf, you’re doing so out of convenience. Can’t really blame you though. Large parts of the internet is a bout to reach terminal stages.
As you’re using Duck.ai the privacy side of it is probably as good as its going to get. However, it still has broader ripple effects. Duck.ai is using GPT, which means that OpenAI is getting a signal to keep on developing GPT, violating your privacy elsewhere, and spewing out CO2. The use of LLMs isn’t limited to just privacy matters, since it involves several ethical questions as well.
Side note: AI is a very broad topic, and it covers good and bad stuff. LLMs are a far more specific topic, and it involves some unique downsides absent in most AI applications.
-
on The quickest nope of a cookies popup I've ever seen in c/privacy@lemmy.ml · 16 Sep 2026
We value your privacy so much that we got rid of the disagree button and found some extra partners who also value your privacy a lot. Like, you have no idea how much they’re paying us. It’s just crazy… Ahem, so that’s why we really value your privacy.
-
on What do you do for email? in c/privacy@lemmy.ml · 15 Sep 2026
Infinite aliases are great. I have an address for every company, and that keeps my inbox very clean and tidy. If one of them turns out to sell my address to spammers, I will know exactly who to blame. It’s also trivial to set up a rule for deleting everything that is sent to that address.
-
on Google’s Emissions Climbed 48% Since 2019. AI Is Why. in c/technology@beehaw.org · 13 Sep 2026
You do have renewable energy. Why not use it?
-
on Dating Apps Are Dying in c/technology@beehaw.org · 13 Sep 2026
Nothing of value was lost.
-
on I am asked for Gmail in c/privacy@lemmy.ml · 12 Sep 2026
Those are some important principles. Recently, I’ve been thinking about following this method in a completely different context.
Using a privacy-respecting phone is entirely doable as long as you don’t need to interact with certain parts of the modern world. In 2010, such problems didn’t exist, but mobile Linux wasn’t available for this kind of use. Nowadays it is, but the world around it has moved on, and what’s required for a daily driver of a normie has changed radically.
So here’s how to adapt: Get some basic Android phone and use it for doing the few necessities that don’t work on a FOSS-based phone. That Android is just a skeleton. There’s nothing too personal, and you don’t use it for anything unless absolutely necessary. Same as that Gmail idea, but adapted.
-
on Watch out: Apple timepiece can grab snippets of conversation without both speakers' consent in c/privacy@lemmy.ml · 12 Sep 2026
There are indeed many eerie similarities. Both books got many themes correct, even though the exact details were wrong. BNW isn’t as popular and well known as 1984, and that’s why the latter still gets all the attention. If people knew BNW, they would be referencing it all the time, since modern life is pretty messed up even by BNW standards.
-
on Watch out: Apple timepiece can grab snippets of conversation without both speakers' consent in c/privacy@lemmy.ml · 12 Sep 2026
Currently we live in a world where neither book became completely true. However, bits and pieces from both did. Nowadays there’s all sorts of enjoyment and distraction easily available, and it seems to partially take the role soma has in BNW.
-
on Grandma finds the secret knitting app in c/lemmybewholesome@lemmy.world · 8 Sep 2026
Wait until you hear about Inkscape. Next level knitting patterns are just around the corner.
-
on There's No Reason to Buy a New Phone Anymore in c/technology@beehaw.org · 21 Aug 2026
So you run Lineage or Graphene on it? The default Android has been without security updates for years by this point, as Samsung just loves to create e-waste to increase shareholder value.
-
on There's No Reason to Buy a New Phone Anymore in c/technology@beehaw.org · 21 Aug 2026
The trick is to buy a used phone every year or two. It will be new enough and the previous owner took the biggest depreciation hit on your behalf. Personally, I like to stay two years behind, because it’s cheaper that way.
Slop grande! That’s a wonderful expression. I’ve been on the receiving end of one quite recently, and I think I’ll start using that word from now on.