Interestingly, LLMs are horrible at Zork: https://arxiv.org/abs/2602.15867

Our results reveal that all tested models achieve less than 10% completion on average, with even the best-performing model (Claude Opus 4.5) reaching only approximately 75 out of 350 possible points

Congrats, you figured out the worst possible title for the post.

Microsoft has over 200k employees, and the people (person?) managing the discord are probably at the bottom of the totem pole. I doubt they put more than 5 seconds of thought into this.

I took thermodynamics as an elective, without realizing that was the weed out class for mechanical engineering majors. Biggest mistake of my life.

Google is alleging the first known version is 7 years old, which would predate AI - but I cannot open the link it only takes me to the community homepage.

There were definitely AI models for restoring images 7 years ago. I remember using https://github.com/jantic/DeOldify (which started in 2018) about 6 years ago to colorize an old photo album, and it worked well.

yep, it’s wood

Is it 1/4lbs with or without the pregnancy test?

You can read it for free! https://archive.org/details/cu31924003102971/page/n5/mode/1up

CSV existed for over 30 years before RFC 4180. Excel, and countless other tools, have their own incompatible variants. Excel in particular is infamous for mangling separators when exporting to CSV.

You just need more practice.

What about the M-expression version (f[x])?

Little tech? Like, a micro company that makes software? A “micro-soft”, if you will.

I work in big tech and this is my life. I envy anyone who thinks you’re exaggerating, because that means they haven’t experienced the joy of spending weeks trying to track down the team responsible for a bug and then months hassling them to fix it.

But how would they rake in profits? These people are only paying $200/month at most, and OpenAI has already said they’re losing money on Pro subscribers due to the cost of running those models. I don’t see how OpenAI is going to become profitable off the consumer market alone, which means they have to focus on the business market. These AI addicts are a dangerous distraction to them, because not only are they losing money the more that susbscribe, but they’re also getting bad publicity and attracting attention from regulators. That could scare off their actual cash cows (enterprise companies).

It changed my life. No words can possibly describe it.

Followed quickly by “vibe bankruptcy proceedings”

Where are the Snowdens of yesteryear?

What is that from?

And it’s highly effective!

Predation accounts for a relatively low rate of nest failure: only 34% compared to an average of 80% for birds in similar habitats. This may be enabled by their well camouflaged nests, or simply the lack of local predators.

https://en.m.wikipedia.org/wiki/White-browed_tit-warbler

Only for male fashion. For the ladies, brown is in.