You might be interested in this (it was super timely for me 6 months ago) : https://leehanchung.github.io/blogs/2026/04/05/the-ai-great-leap-forward/
clif
Just a geek, finding my way in the fediverse.
- 0 Posts
- 9 Comments
clif@lemmy.worldtoData is Beautiful@lemmy.world•Acceptance of homosexuality across the USEnglish
1·10 days agoI was pretty disappointed to see Arkansas as the worst… Usually we can rely on Mississippi or Alabama to keep us from last place on anything : (
clif@lemmy.worldto
Ask Lemmy@lemmy.world•Does anyone here use anything other than Discord for chatting? What exactly and how much?
1·10 days agoIRC is still relevant… at least for the olds.
Most of my channels are VERY low traffic, just occasionally catching up with people I met there 20+ years ago. But, the #xkcd (Slashnet) one is pretty active if you want to jump back in and actually talk to people.
Bouncer on the home server, hexchat on the desktop. I had some mobile client but I didn’t use it frequently enough to keep around.
clif@lemmy.worldto
Technology@lemmy.world•Apple announces touchscreen MacBook Pro with OLED displayEnglish
6·10 days agoSomeone touching my perfectly clean screen and leaving a fingerprint deeply angers me in a way that almost nothing else does. I don’t know why but it’s infuriating.
Yes… But also working on your vocabulary!
… So double yes
Heard this trick on the radio a few months ago and it seems to work for me.
Think of a word, I usually do something short like “cat” or “dog”. Let’s use “dog” for this example.
Take the first letter and list every single word you can think of that starts with the same letter. So, for “dog”, think of all words that start with the letter “d”
Once you run out of “d” words or are taking more than a few seconds between each one, go to the next letter and do the same thing. So, all words that start with “o”. Continue until you’re asleep or have to pick a new word.
I was ready to type “just fire QA and have the LLM do it” then I read the last paragraph.
I’m hoping this is the line that executives won’t cross
Hate to be a downer here buuuut…
clif@lemmy.worldto
Technology@lemmy.world•‘Doom Loop’: OpenAI and Microsoft Admits LLMs Are Destroying the Web and Built on TheftEnglish
0·2 months agoBecause it was generated by a llm.
I did a search for the torque spec on a castle nut last week and the second result was for a nut (as in, food nuts that you eat) website and the llm had gone all in on nuts and added a page about castle nuts. It even generated an image of a castle for the top because… castle nut.
It proceeded to provide vague instructions and a torque recommendation of 2x the actual spec.
Then there was the other one about a small 4 stroke engine where it stated “other sites will say you don’t need to mix oil with gas but you absolutely must!” (You don’t and shouldn’t) … I wonder how many people have fucked up their shit by trusting llm garbage without knowing better.

It makes me really, really happy that I know this insult. Might be the first time I’ve ever seen it used in text (at least text in a language that I can understand)