SHP Forums

I see they have an Odin colored Husky...

Tawmis wrote: ↑Mon Jul 13, 2026 11:48 pm I see they have an Odin colored Husky...

... and Ken has a Quest for Glory 4 T-shirt!

Seriously, though, I read Ken's entire post, especially about AI. I had a seminar this morning about "how to use Copilot". Copilot is probably one of the simplest AI models there are; it's not very powerful, at least not as far as I'm concerned. Most of that is because I've only used it for Excel, and it's geared towards beginners, whereas I've been using (and programming) Excel ever since Excel v. 3, back in the dinosaur days of Windows 3.1 (when it ran under DOS).

Now, you tell some of the young 'uns today about DOS, and they'd be like "Huh?" Never mind telling them about ASCII or 16-bit EGA or reprogramming BIOS. They'd stare at me like a cat stares at a human doing a card trick.

I know AI is coming. I know it's the future. I also know a lot of people think that it's the bee's underpants and the greatest thing since sliced kumquats, and to to those people I have a one-word answer: Skynet.

No, I'm not an alarmist, and I don't believe AI will try to "rise up" and kill us all. But the people writing the AI probably want to take over at least a part of what we do and who we are. And to them I give a big raspberry.

Rath Darkblade wrote: ↑Tue Jul 14, 2026 8:53 am

Tawmis wrote: ↑Mon Jul 13, 2026 11:48 pm I see they have an Odin colored Husky...

... and Ken has a Quest for Glory 4 T-shirt!

Seriously, though, I read Ken's entire post, especially about AI. I had a seminar this morning about "how to use Copilot". Copilot is probably one of the simplest AI models there are; it's not very powerful, at least not as far as I'm concerned. Most of that is because I've only used it for Excel, and it's geared towards beginners, whereas I've been using (and programming) Excel ever since Excel v. 3, back in the dinosaur days of Windows 3.1 (when it ran under DOS).

Now, you tell some of the young 'uns today about DOS, and they'd be like "Huh?" Never mind telling them about ASCII or 16-bit EGA or reprogramming BIOS. They'd stare at me like a cat stares at a human doing a card trick.

I know AI is coming. I know it's the future. I also know a lot of people think that it's the bee's underpants and the greatest thing since sliced kumquats, and to to those people I have a one-word answer: Skynet.

No, I'm not an alarmist, and I don't believe AI will try to "rise up" and kill us all. But the people writing the AI probably want to take over at least a part of what we do and who we are. And to them I give a big raspberry.

To me, Ken's very blog - and his blindness to it (or his lack of caring about it) - is the very problem with AI.

He literally made a "game/simulator" with never - not once - looking at the code.

People, like Ken, will do just that. They will use AI to help them generate code/fix code at first, but soon enough, they will have AI program the entire thing, start to end, never looking at the code.

And what's going to happen when servers go down? And there's a problem in the code?

Who is going to fix it? If no one is looking at the code?

The code may be insanely overly complicated.

And Lord forbid, for those who rely on AI to do all the code, they'd have no idea, in time how to fix the code.

That's my concern. You can use AI to tell you a book in a paragraph summary. You can use AI to even WRITE a book for you. AI can now create MUSIC, with bands, instruments, vocals.

At first, AI was a "threat" to the creative folks (writing, art, music) - but it's clearly spread into now job markets as more and more lay offs continue to happen.

The world is going to be a sad state, when there's no jobs because of AI, people are unemployed, and can't afford computers any more because of the insane costs of building data centers to support AI.

The more I read Ken's blog, the more I saw he was further and further detached from the whole Sierra family type dream he had originally created.

And that, makes me sad.

Tawmis wrote: ↑Wed Jul 15, 2026 12:51 am

Rath Darkblade wrote: ↑Tue Jul 14, 2026 8:53 am

Tawmis wrote: ↑Mon Jul 13, 2026 11:48 pm I see they have an Odin colored Husky...

... and Ken has a Quest for Glory 4 T-shirt!

Seriously, though, I read Ken's entire post, especially about AI. I had a seminar this morning about "how to use Copilot". Copilot is probably one of the simplest AI models there are; it's not very powerful, at least not as far as I'm concerned. Most of that is because I've only used it for Excel, and it's geared towards beginners, whereas I've been using (and programming) Excel ever since Excel v. 3, back in the dinosaur days of Windows 3.1 (when it ran under DOS).

Now, you tell some of the young 'uns today about DOS, and they'd be like "Huh?" Never mind telling them about ASCII or 16-bit EGA or reprogramming BIOS. They'd stare at me like a cat stares at a human doing a card trick.

I know AI is coming. I know it's the future. I also know a lot of people think that it's the bee's underpants and the greatest thing since sliced kumquats, and to to those people I have a one-word answer: Skynet.

No, I'm not an alarmist, and I don't believe AI will try to "rise up" and kill us all. But the people writing the AI probably want to take over at least a part of what we do and who we are. And to them I give a big raspberry.

To me, Ken's very blog - and his blindness to it (or his lack of caring about it) - is the very problem with AI.

He literally made a "game/simulator" with never - not once - looking at the code.

People, like Ken, will do just that. They will use AI to help them generate code/fix code at first, but soon enough, they will have AI program the entire thing, start to end, never looking at the code.

And what's going to happen when servers go down? And there's a problem in the code?

Who is going to fix it? If no one is looking at the code?

AI / the Terminators will fix it, of course!

Tawmis wrote: ↑Wed Jul 15, 2026 12:51 am The code may be insanely overly complicated.

And Lord forbid, for those who rely on AI to do all the code, they'd have no idea, in time how to fix the code.

That's my concern. You can use AI to tell you a book in a paragraph summary. You can use AI to even WRITE a book for you. AI can now create MUSIC, with bands, instruments, vocals.

Seriously, I agree. I've been programming all my life. Now AI comes along and gives just anyone the "ability" to "program" - but not the ability to debug your program.

As you say, if something goes wrong - and it will, because the more things you want your program to do, the more complicated the code becomes - they have no idea how to fix it.

I also foresee AI creating "spaghetti code", i.e. over-complicated and obtuse code that's impossible to unravel, like a big ball of spaghetti.

Tawmis wrote: ↑Wed Jul 15, 2026 12:51 am At first, AI was a "threat" to the creative folks (writing, art, music) - but it's clearly spread into now job markets as more and more lay offs continue to happen.

The world is going to be a sad state, when there's no jobs because of AI, people are unemployed, and can't afford computers any more because of the insane costs of building data centers to support AI.

The more I read Ken's blog, the more I saw he was further and further detached from the whole Sierra family type dream he had originally created.

And that, makes me sad.

AI is still a threat to creative folks, no double-quotes needed. We've seen it happen with both creative folks and programmers in the NineZyme / Microsoft situation: Microsoft is laying people off and focusing more and more on AI.

It reminds me very much of the goose that lay the golden egg.

Ken is retired, so he can do what he likes. Obviously boating is now a higher priority for him than making games, so if AI is more economical for him than hiring actual people, so be it. Yes, it's sad, but what Ken does is up to him.

I think that as AI becomes more and more pervasive (or should I say invasive?), we'll see more and more people making games. Steam is already experiencing a glut of indie games. I've seen some people say that's good for consumers -- more games mean more choice and lower prices, hooray -- but I don't think so.

A-list studios like Bethesda, Microsoft, Blizzard, Firaxis and Larian will continue to charge top dollar. Indie developers like Ken & Roberta, the Coles, and NineZyme won't be able to, so they'll need to sell more units to compete. (Of course, they're not competing with the likes of Microsoft, so that point is rather moot - but they'll need to sell more units to recover their investment).

Not only that, but more games will mean a much harder time for gamers like us to sort the wheat from the chaff, the good from the bad, the cream from the crap.

And that makes me sad.

I thought Ken's blog post was measured. ("I’m not sure whether I should be excited about the future or horrified by it. My suspicion is 'horrified,' but opinions differ." "That is both exciting and terrifying.") He wanted to see what AI could realistically do and the only was to find the answer was to try. Sticking one's head in the sand will not make the problem go away. I personally don't use it for work and I've mostly just toyed around with it (although I did delete my ChatGPT account after they got involved with the Pentagon): trying to get it to hallucinate, having it write short stories, and just talking with it. I'm not a programmer, but I did take some programming classes in college and it could do my homework assignments in seconds. I'm not sure how capable AI is with high level coding, but I do know that the people working on DOSBox Staging are using it to review code. Like it or not, AI is not going away. I think the best comment I read about this was "AI writing poetry while people do manual labor is a future no one asked for."

And verifying the code is correct is a huge concern. There are all sorts of security issues with AI too like prompt injections. There's a webpage where AI agents "talk" to each other in a forum. These agents have access to their users information, and you just have to hope that they don't divulge anything private. Instructions are treated like suggestions. You just have to hope they follow them. In some ways it reminds me of the Internet-of-things. Why would you want an oven connected to the Internet? So you can pre-heat the oven while you're away via an app? That creates a huge security risk for a minor convenience.

And of course there is the massive amount of power and water that data centers demand. And putting them into space is just stupid.