Here's an expensive thing!
What value does it have?
...you figure it out!
I am not impressed.
:o
Deestan
joined 2 years ago
The lemmyverse isn't huge, but broad.
Unless you give some examples, people are probably going to respond based on what they saw somewhere else than where you were thinking of.
My best attempt at an answer is that lemmy has fewer long-term well behaving residents, so people coming from other platforms because they kept being banned for bad behavior or just kept not getting along, stick out more.
Sir, this is a Wendy's
This reads like OpenAI's fanfic on what happened, retconning decisions they didn't make, things they didn't (couldn't!) do, and thought that didn't occur to them. All indicating that the possibility to be infinitely better is not only possible, but is right there for their taking.
For the one in April, engineers created many new versions of GPT-4o — all with slightly different recipes to make it better at science, coding and fuzzier traits, like intuition.
Citation needed.
OpenAI did not already have this test. An OpenAI competitor, Anthropic, the maker of Claude, had developed an evaluation for sycophancy
This reality does not exist: Claude is trying to lick my ass clean every time I ask it a simple question, and while sycophantic language can be toned down, the behavior of coming up with a believable positive answer for whatever the user has, is the foundational core of LLMs.
“We wanted to make sure the changes we shipped were endorsed by mental health experts,” Mr. Heidecke said.
As soon as they found experts who were willing to say something else than "don't make a chatbot". They now have a sycophantically motivated system with an ever growing list of sticky notes on its desk: "if sleep deprivation then alarm", "if suicide then alarm", "if ending life then alarm", "if stop living then alarm", hoping to have enough to catch the most obvious attempts.
The same M.I.T. lab that did the earlier study with OpenAI also found that the new model was significantly improved during conversations mimicking mental health crises.
The study was basically rigged: it used 18 known and identified crises chat logs from ChatGPT - meaning the set of stuff OpenAI just had hard coded "plz alarm" for, and thousands of "simulated mental health crises" generated by FUCKING LLMs meaning they only test if ChatGPT can identify mental health problems in texts where it had written its own understanding of what meantal health crisis looked like. For fucks sake of course it did perfectly in guessing its own card.
TLDR; bullshit damage control
I am scared of the "TRY ME" button in front of the smiling, plier-wielding elf
Popping a pimple
Business idiots are killings jobs. Generative AI is just their excuse to do it and threat to make people feel more replacable.
It's on the verge of pedantic, but I feel it important that we blame people for lying and causing harm, and not let them hide behind the imagined inevitability of tech and progress.
Generative AI can't replace shit, but the lie that they can and do, is the weapon wielded more than the tech itself.
This platform would have to have all the same functions
This expectation comes from inertia, not need. No system, thing, or product can ever succeed at checking everyone's little boxes from another product in a satisfactory way.
Also both you and your friends are older, different people now. The old magic will not come back. Figure out what you actually actually need and find something new that will be good at that.
Facebook was new and confusing once. If you can face that again, you'll find beautiful things.
A $10 charity donation in his name
Take at least three old socks, fill them each with a fistful of dried peas, lentils or beans. Sew them shut and cut off the empty part. The end result should be roughly ball shaped soft objects that fit in your hand.
Now, spend at least two hours every day practicing juggling them. Start with one. While staring straight ahead throw it in an arc from left hand to right so that it passes in front of your face. Use circular motions.
The daily physical movement will do wonders for your mood, but most importantly: In a few months you are going to be impressively good at something cool that people around you suck at.
Self-esteem from skill and personal development is more healthy and sustainable than... this.
I always heard their little quips as "My wife for Aiur!"
R U L E
Most of them can be solved with various applications of patience, so the trickiest one was the Express Delivery - winning Space Age in less than 40 hours.
Some tips for others who want to try it:
You can not reduce biter base sizes without invalidating the achievement, but you can turn down pollution, biter evolution and biter expansion. Turning the starting area size up to max means you can get all the way to Aquilo without seeing a single biter.
Resource patches can be turned up high and cliffs can be turned down low.
The scale needed is absurd and the main strategy should be on being able to scale hard and fast. My Nauvis base had 20 rocket silos going constantly, two 4-reactor nuclear plants, and it still felt a bit underscaled when I got to Aquilo.
Nauvis was the easiest to get to bigscale in this timeframe, as it already had a head start by the time the other planets were operational. Don't bother setting up blue chips on Vulcanus. Copy-paste the blue chip production on Nauvis and just ship them over with rockets.
What worked for me to scale:
- BOTS BOTS BOTS thousands of bots immediately and forever on all planets.
- Make sure to produce at least 2 per second of any science.
- Nauvis is the backbone. It can just ferry nuclear fuel, blue chips, rocket fuel, and low density structures to all other planets. (Fulgora gets trivially self-sufficient after a while though)
- Make production modular. Not necessary with perfectly symmetrical or cityblock, but enough that you can just zoom out, copy-and-paste "the area that turns iron ore, copper ore, and raw oil into blue chips", and just connect it to some new miners or a new oilfield.
- Inefficiency and waste are better than starved belts or constant tuning problems. Need 20 red chips per minute? Copy the thing that makes 20 red chips per second. Give it a separate set of miners, and don't care if it spends most of the game idling.
- On each planet, establish a bot network and stable power, then GTFO to the next one. Build the rest via remote.
My time was 36 hours, following these rough milestones:
- Logistic system before 10 hours
- Aquilo before 30 hours
Update on https://lemmy.world/post/26605581
Playing on modified Marathon setting: 100x science cost instead of 4x
I think I'm finally over the challenging part of the challenge. Until now it was extremely uncertain whether my base would both survive biter evolution on Nauvis and manage to claim enough resources to get to space and Vulcanus for "unlimited" iron and copper.
At some point the biters evolved past my tech: I can't assault behemoth biter bases with only low damage tech red bullets, effectively making any unclaimed resources unavailable until I could tech up (which would need resources).
I can, however, guard against behemoth biters with walls of red bullet turrets and flamethrowers.
Vulcanus also turned out to not be as straightforward as I hoped. While I do get a lot of resources in the starting area, the initial tugsten patches are guarded by small demolishers. Fortunately, 100 turrets of red ammo and spamming poison capsules proved enough to clear out a significant area and secure enough tungsten to get me far up the tech tree.
I feel like I "won" the challenge now that I can safely tech past any future hurdles, and the remaining tasks are just putting in the hours. Not sure if I'll keep playing past this, but it was a ton of fun. :)
For anyone who wants to see a true maniac attempt what I assumed to be impossible: Michael Hendriks may be able to win with 1000x science cost, exploiting deep knowledge of the biter expansion algorithm, spending tens of hours filling the map with pipes, and meticulous calculation of resource usage through the tech tree carefully adapting or not adapting quality, modules etc.
Some screenshos of the current state:
They were stress-bored and were fidgeting by tapping their watch so much that it accidentally triggered the emergency mode and sent me this SMS.
Alert not intended but also in some way accurate.
From the comic "Girl Genius" by Phil and Kaja Foglio
Just for fun!
Made a shape out of gray self-drying hobby clay, took a (very clumsy) silicon mold of it, and now I have a fun shape to pour excess soap into if I make too much for the main mold.
On the left: clay thing. On the right: lavender soap.
view more: next ›
I have briefly scanned the headline and am ready to share my immediate opinion on this significant political issue that I did zero research on. Where do I put it?