rossirpaulo

This is great! We had a similar thought and couldn't agree more with "LLMs prefer producing something rather than nothing." We have been consistently requesting responses in JSON format, which, despite its numerous advantages, sometimes imposes an obligation for an output even if it shouldn't. This frequently results in hallucinations. Encouraging NULL returns, for example, is a great way to deal with that.

That's an interesting point on Authorization. Though, if Google does it, there must be a way for you to serialize the data too. Curious to know more as well.

Anyways, excited to see where this goes! I've been reaching out to people for a consumer-oriented idea, and forums have been a great source of info; making that search more specific and succinct is much needed. This would really help me out!

That's completely right. Fun fact: We're over 70,000 new indie titles per year already!

That's a great question! Initially, MagnaPlay's entire idea was building a smaller portfolio (~50-100 titles) compared to our competitors; we'd be focused on quality over quantity -- essentially becoming the place to play the best indies. We would do that by collecting player engagement data, serializing reviews, and flat-out asking for which games our audience wanted to see on our platform. While we still haven't changed that strategy in the short term, we see ourselves adding more and more titles with time and eventually allowing MagnaPlay's community do all the curation instead of relying on more abstract agents like the team's preferences :)

So, picture this: if you like puzzle games, for instance, you would essentially be put in the "Librarian" role/house (much like how Hogwarts would put you in Gryffindor if you're brave). All so you can see fewer and fewer titles that aren't worth your time!

Interesting idea! Adding MagnaPlay to the Steam Deck might pose an exciting challenge... We've developed our platform on a cross-platform stack, so it might be possible. I'll take a look at the feasibility here.

And hey, indies are super casual, and a 300Mb launcher runs on any system. Give our platform a chance; maybe it is just casual enough for an evening gaming session after you close down all your Chrome tabs and spot our logo on your Desktop...

That's right; I agree with you on the content creation point. We do plan to make originals, much like Netflix creates their own TV shows and films. Yet that's quite a burn for an early-stage startup like ours. The long-term vision of MagnaPlay must encompass funding developers and adding their games to our portfolio. However, in the meantime, all we can do is upstage the indie segment by giving them visibility, staying true to our cause, and supporting developers with community-friendly features. For instance, players can distribute 10% of their subscription to any developer, essentially crowdfunding exciting teams and projects.

Makes total sense. We talked to other people who had a similar headspace: exploration is essential, though purchasing each game is not necessarily the problem. One of the things you gain from subscription service is curation; the idea around there is to minimize those encounters where you absolutely dislike the game, either because there's an audience voting for the titles you see or because there's a recommendation system that filters stuff you might not like.

On the point of trying out more stuff before you buy them: we considered adding a section on the platform for alpha/beta releases. It would allow developers to test and get feedback throughout the early stages of development while allowing players to engage with titles they would otherwise avoid. It would be separate from the revenue share model as these games wouldn't be fully baked and, therefore, at a lower standard than those on the "official" portfolio. We might be able to roll out this feature in the coming weeks! Would that be something you see yourself using?

And absolutely, in the context you gave, you would help developers a lot with our model. Think about it this way: the 200 hours you played on that game got compensated once: when you bought it. On the other hand, if you are subscribed to MagnaPlay and mostly play that game, you'd be essentially giving 80% of your subscription to its developer every month.

That's a good idea. We like to think that subscription services foments this idea of exploration; after all, it makes the traditional process of finding, purchasing, installing, and finally playing shorter and smoother.

Excited to be building this with you, Pedro!