It has been a pretty quiet week where I haven't been doing any work on my text adventure for NaNoWriMo. Actually I've been feeling the pull towards other project ideas that I'd really like to get onto once November is over.
Read
Alas, my reading has mostly been of social media which always evaporates from my mind as soon as I read it. I do recall that there were quite a lot of things being reported and talked about this week such as various software issues at companies, launch of GTA6 and Anthropic's IPO.
One thing I did rather like was this article by Amos on Terminal Levy. The website is a place for musings on games and game design. The latest article discusses the pros of limited graphics allowing you to conjure up more memorable vistas in your mind and the issues of identifying game objects in hyper realistic surroundings.
It seems silly to mention but I did read through an art book called 15 Minute Art: Pen and Ink by Anna Tromop. I mostly picked it up because the art exercises in it looked both nice and achievable. Of course I haven't actually done any of the exercises in it yet, only read through the instructions.
Listened
I listened/watched a different podcast show this week, an interview with Paul Christiano on the Dwarkesh Patel Podcast. Dwarkesh has become a pretty big name in the tech space having interviewed many different researchers in AI. This episode was with the head of the Alignment Research Center who is also a member of the board of OpenAI.
It was a pretty sobering interview with Christiano saying that he expects AI progress to be extremely fast from this point, a matter of 6 to 12 months until we have superhuman AI models. He is very concerned that they'll reach a level where they'll be able to disrupt governments and doesn't feel like we are on track to having them aligned by then.
Indeed the specter of disruption due to AI systems in the hands of bad actors or running loose on their own seems to be hanging over a lot of things at the moment. In the UK the government is actually telling us to have stockpiles of food and water in case of serious disruptions.
Christiano, speaking as a private citizen, is very pro AI labs being a lot more transparent about their alignment research and for them to completely stop researching algorithmic improvements. He also is very pessimistic about labs being able to stop the weights of their models from being stolen meaning he thinks China will have dangerous superhuman AI at about the same time as the US.
Watched
I saw a vtuber, Margo, play through some of an old Indiana Jones game that I used to play called Indiana Jones and the Infernal Machine and boy, did the game look bad now. To be fair she was playing the Nintendo 64 version rather than PC so the graphics were even more compressed and glitchy but the game design was also rubbish. The movement controls were so clunky and the level design really difficult to parse. It is nostalgic to see someone play through it but it looks like pure suffering.
Another vtuber I sometimes watch is Holly and she was playing through Tower of Gold by Dan Remar. Remar is an old name in indie game development having made the classic Iji. All his games are freeware and his latest is no different. It has this lovely Amiga-like artstyle with only 16 colours. The characters in the world are interesting and the gameplay is much more about exploration rather than combat.
Of course why am I watching someone play it when I could be playing it myself. I guess I enjoy sharing the experience of playing a game with someone else, even if that is vicariously. Still I'd strongly recommend people checking out the game themselves here.
Drew
I went along to Art Class this weekend and did a drawing of a shiny letter balloon. It was nice to have such a simple study in monochrome colours. I'm reasonably happy with how it came out and it was well received by the rest of the class. I do like art when I don't spend too long on it and it comes out well.
Some other art things I did was looking at art tutorials by Pikat which were well explained and watched a video retrospective of someone learning to draw over a whole year here. Both were in an anime style which I'd like to be able to replicate myself.
The second video was quite confidence boosting in terms of how bad their art was to begin with but they kept at it and look at how much they improved. I would love to have their confidence in sharing art which seems so bad as a way to gauge improvement.
Game Design
Bit of an odd category but bear with me. Instead of listing off more games that I watched footage of I thought I would go over the general category of them at once.
There seems to be something of a booming genre of what could be called labour intensive games. The first that springs to mind is Find the Needle which is about sorting through a haystack to find one needle but there are lots of games which start you out by doing things laboriously before letting you automate them.
Possibly the inspiration was Factorio as many of these games involve building up automated processing lines to get through the massive amount of material you need to get through. The important thing seems to be the scale of the task and that even with upgraded tools you are still chipping away at it.
Often friends can join to work together at the task such as in How They Built the Pyramids or Moon Sweepers. Things tend to be more manageable when done with friends and it can be nice to hang out and do some low intensity activity together.
Other ones involve truly huge amounts of things to collect like Quadrillion Carrots which seem to be reifying the absurd numbers that you would be getting in the likes of Cookie Clicker. I think it is a praiseworthy goal to make games where you really grapple with how massive some quantities are to give you appreciation of thousands, millions, even billions of things.
Another aspect of these games that I like is how they model exponential growth which can be unintuitive to get your head around. Say in Moon Sweepers it might take hours to get the first 0.01% of the whole moon clean but as you clean it you get more money to spend on better tools and automation so that percentage completed can balloon up.
It is a really cool use of modern computing power to have games which grapple with such large numbers of objects and truly massive growth rates as it isn't something that could be done in such an intuitive manner before. The modern world involves such vast amounts of things so it helps for people to experience them in as high fidelity as possible. I guess that means lots of games that seem to involve doing very repetitive tasks to begin with. They're fun, honest!
Other
As I mentioned before I didn't make any progress on my text adventure and only wrote about 10K words in my journal. It's a shame to say this but I think it is an example of how I really struggle to stay on task for more than about a week of concerted effort on something.
Indeed I have felt my mind drift to other things that I'd like to be doing this week such as art and game design. I'm not sure how the rest of November will go. Perhaps it would be best to follow my interests in the moment rather than doggedly sticking to saying I'll do one thing if I don't even end up doing that one thing. Being self motivated is hard.

No comments:
Post a Comment