Comments 10

Re: ChatGPT Translated An Article About Space Harrier, Then Suggested "Tailoring" It For Retro Gamer

Nischenliebhaber

This is a much better article than your last one on the topic and is of the standard I have come to expect from Time Extension, thank you.

I think that most media outlets are already using AI. Take Gamestar here in Germany, for example: a large website that has been generating hard clickbait for many years. No journalism whatsoever. Of course they use AI or will use, because they have been training their audience to consume this content for years. Whether it comes from a human or AI is no longer noticeable.

And it's the same with many other tabloid/clickbait media outlets.

And here's a good article from the NY Times on how AI could displace entry-level jobs: https://www.nytimes.com/2025/05/30/technology/ai-jobs-college-graduates.html?unlocked_article_code=1.PE8.Z1uC.IKjWNUrtCXPz&smid=nytcore-android-share

Re: OpenAI's ChatGPT Lost A Game Of Chess Against The 48-Year-Old Atari 2600

Nischenliebhaber

@Damo

Here in Germany, many teachers no longer give homework assignments because of AI. Or IF they do, then they give assignments that really can't be done easily with AI – such as working together to create posters. In other words, creative tasks where AI can only provide support at best.

Of course, it's dangerous to blindly trust AI when politicians here in Germany suddenly start talking about letting AI make decisions about social benefits, without realising that AI has no place in these areas. That should be completely rejected. We're not that far apart on this. However, I would like to see more public discussion of such important ethical issues. Okay, this site isn't the right place for that – I know. ^^

On the other hand, I always find it inspiring to see the possibilities AI offers people. For example, someone used AI to teach themselves how to program text adventures and developed a randomised game with puzzles. The texts and puzzles are all his own creation (the gameplay is randomized), but the creator taught himself programming with the help of AI. I think it's cool that you can use it in an empowering way to realise an old childhood dream (the instructions are really worth reading!): https://vanterraarts.itch.io/gold-miner

Re: OpenAI's ChatGPT Lost A Game Of Chess Against The 48-Year-Old Atari 2600

Nischenliebhaber

@Damo
That's not the point. I myself am very critical of AI. BUT: You have to remain objective. Just because some guy on the internet can't handle AI doesn't make for a good story. There are so many real problems caused by AI. And yes, that does include the loss of jobs, for example in graphic design or IT. Or the fact that video games are becoming increasingly generic due to AI, etc. But articles like this make AI critics look more like gamergaters.

I mean, at the end of the day, AI is a tool. One that saves a lot of time, for example in the academic transcription of interviews. Or in the creation of shift schedules, which would otherwise take you at least six hours – with a good reasoning model, it can be done in 30 minutes.

You sounds like the people here in Germany 25 years ago who saw the internet as an “overhyped gimmick” that wouldn't catch on and that there was no need to prepare for a digital society. The consequences are still being felt today.

Re: OpenAI's ChatGPT Lost A Game Of Chess Against The 48-Year-Old Atari 2600

Nischenliebhaber

@Daniel36
The explanation for the problems with your AI adventure is quite simple: AI models are not very good at remembering chat histories. This means that the longer the chat, the more the AI stutters. This is the case with every provider: ChatGPT, Gemini, Le Chat, etc.

"But asking it to fact check something, isn't the best thing to do." It's all about the prompt. But a good alternative is Perplexity! Best search engine for me. Because traditional search engine are absolutely NOT useable with all those SEO-spam, fake news, etc.

About this “news”: Clickbaiting. 4o is NOT designed for such tasks. It is also not a reasoning model - i.e., a model that first searches for errors before outputting text. With reasoning models, this would not have happened. But hey, you can generate clicks and jump on the “AI is sooo stupid” bandwagon. Do you really need that?

Re: 20 Years In The Making, Mario Adventure 3 Is The Ultimate Mario 3 ROM Hack

Nischenliebhaber

@Boston_J This has the technical background that this area was left free in order to save storage space, as it was covered by the CRT televisions of the time anyway. Many NES and Master System games have this unused space of 8 pixels to the left and right, sometimes also at the top and bottom. In emulators such as Mesen, it is therefore possible to cut off this “overscan area”.