That’s a neat experiment for several reasons. It shows limits of LLM capabilities, the importance of training data, context sensitivity, very dramatically shows that LLMs should not be trusted with important tasks if not supervised and that their advice has to be taken critically.
The link to their source is boldfaced and underlined within the article. You have missed it as it seems:
Here it is:
https://www.emergence.ai/blog/emergence-world-a-laboratory-for-evaluating-long-horizon-agent-autonomy
I still haven’t played through Skyrim.
Apparently you’re right. I misunderstood it a bit.
Trü.
I guess I stumbled upon the wording “how awful being fat is on your body”.
That’s why rich people are bad people. If you are at the capacity to easily mitigate suffering in the world, but won’t do it for personal enrichment, you’re an asshole.
People using Meta’s products are masochists.
“Yes, daddy, please violate my privacy even more and manipulate me into hating myself and making me addicted to your shit.”
You do not need to look a certain way to be worthy of respect, kindness, and love. And it’s sick that so many people are such superficial assholes.
A healthy body needs fat. As long as you are healthy, you can have fat. The amount and thresholds can vary strongly from person to person.
Oh no, someone used a big babble machine for brainstorming ideas! How despicable! /i
That’s how freemium works: fIrst get you hooked, then make you pay.
That’s the neat thing: never, because they explicitly tell you that big babble machines make mistakes and you’re responsible for any actions you take on that.
The important question is, whether the ex made lemonade.
Looks like they are screwed.
In fact, you can.
How good it will be, how performant and how fast you’ll have it ready is an entirely different question.
There are plenty of open source models though that can be run locally. So getting a beefy server and running a local LLM there might already do sobe of the tasks you need the big babble machines for.
Devs cannot just use a different platform without cutting out a huge userbase.
Most gamers I know, don’t care that much about where they get their games from. Heck, they would buy it from a dubious guy with USB sticks who is travelling on a donkey, if they had to.
Make a good game and people will follow.
I suppose the issue is rather about discoverability and getting people to learn about a game they might like. But steam is not necessarily easier in that regard. You’d still need to do some marketing to get the word around. If you’re lucky by scratching the itch of a lot of people, steam can boost sales by its trending and discovery lists, sales highlights, etc. But that’s not easy to achieve just by choosing steam as a distribution platform.
That’s one of the issues I meant.
Ahem…
https://en.wikipedia.org/wiki/Night_soil
We need to address a couple of issues first before making this a thing again, but in principle I think it’s a worthwhile direction.
There is quite a large chunk of professors who need the interaction with students and see their faces and reactions to be happy on the job. That was most apparent when covid shit hit the fan and the time after when they were all so incredibly happy to see students onsite again (and no longer recorded classes unfortunately).