mid_gen's Launch a Game Studio Thread

[mid_gen]

Finally got another vlog up, ended rambling a fair bit, blame the heat! I couldn't sit in my sweatbox study editing it for too long so it's nearly 30 minutes I'm afraid

Been down to De Montfort Uni twice this week, once for the Games Tech grad show, and yesterday for the Games Art. Some good contacts, lad from Nottingham I'll probably get in to do some game design work if he's interested (I've refactored all the ship/weapon code so all the blueprints need redoing). Also a good concept artist I've exchanged a few messages with, will commission some concept explorations of the visual style.

Now learning Figma to help speed up creating socials content. I've gone public with the studio on LinkedIn now so want to keep that momentum up with posting for the forseeable. Nothing specific I'm aiming for, just visibility at this point, before I get to approaching investors/publishers.

 

[mid_gen]

Still plugging away, and making a lot of good progress on the prototype. I've ended up chatting more on the Discord than posting on here! A recent vlog update :

Also, a monologue/rant/advice video for young programmers about breaking into the games industry :

 

[mid_gen]

Well, I did not get the UK Games Fund grant. Mostly positive feedback which is good.

Frustratingly, the main negative feedback was that my saying that I'm using AI tools to streamline development was 'unproven and risky'.

You have to run this absurd roll of the dice at the moment to try and second guess if the person you're talking to is in the pro or anti-AI tribe....really ridiculous. If you speak to investors all they want to hear is how you're using AI to streamline your production. But applying for grants you gotta pretend you aren't using them. Daft.

Anyway, that was the ideal funding scenario, but there are other paths forward!

 

[100Ryan]

As I understand it these grants are mostly job creation programmes for their country to boost the number of people in a particular industry. From that perspective it is logical that they would prefer to fund something that is creating jobs and work for more people in the UK rather than sending money to US AI companies. The fact that you might accomplish something impressive individually along with AI (and contracting out some work as you've mentioned before) might be outside of their intended purpose?

 

[mid_gen]

As I understand it these grants are mostly job creation programmes for their country to boost the number of people in a particular industry. From that perspective it is logical that they would prefer to fund something that is creating jobs and work for more people in the UK rather than sending money to US AI companies. The fact that you might accomplish something impressive individually along with AI (and contracting out some work as you've mentioned before) might be outside of their intended purpose?

You are right that the funding aims to create jobs. That’s what I would have spent the grant on, hiring freelance artists (which is what it was intended for, and they agreed my hiring plans were solid).

The money I spend on AI is trivial compared to the cost of hiring people, really barely worth considering.

The net result of not getting the grant though is that I can’t afford to hire artists right now. I’m going to have to do it myself, which is far from ideal, but I do have some experience, so not a game over scenario.

I have heard through the grapevine that the people behind the fund are anti-AI. Unfortunately after I did my pitch. Lesson learned.

 

[Noxia]

Well, I did not get the UK Games Fund grant. Mostly positive feedback which is good.

Frustratingly, the main negative feedback was that my saying that I'm using AI tools to streamline development was 'unproven and risky'.

You have to run this absurd roll of the dice at the moment to try and second guess if the person you're talking to is in the pro or anti-AI tribe....really ridiculous. If you speak to investors all they want to hear is how you're using AI to streamline your production. But applying for grants you gotta pretend you aren't using them. Daft.

Anyway, that was the ideal funding scenario, but there are other paths forward!

Edit. Thought it was gov funded thing.

 

[mid_gen]

Edit. Thought it was gov funded thing.

It is government money, comes via a non-profit, but it is taxpayer money.

 

[Noxia]

It is government money, comes via a non-profit, but it is taxpayer money.

Oh in that case I wrote: "that's a bit rich considering multiple Gov departments pushing for use of AI"

 

[Kenai]

Oh in that case I wrote: "that's a bit rich considering multiple Gov departments pushing for use of AI"

Without descending into politics too much, there is no doubt consideration for the optics of job creation within government departments (vs increasing productivity) as 'wasteful' that is percieved differently if that funding is used for job creation in the wider market to 'drive economic activity'.

 

[Worthy]

It's fascinating to see the progress on this. I really like the art style.

 

[mid_gen]

It's fascinating to see the progress on this. I really like the art style.

Glad it's interesting, I'm certainly learning a lot figuring it out as I go

I've had this little nagging idea in the back of my head for a while now, about making the combat phase into a kinda rythm game, like Rez. Have an electronic soundtrack and have to assign the weapons to different instruments. I might let an agent run off and implement it on a branch and see if it's got legs.

 

[NickK]

Don't forget you could monetise for people to drop in on your development work for a two hour slot. People will pay to be interactive with someone actually doing AI development to understand more - specifically more senior people that want less BS/slideware and more programming alongside. Takes time to vet and arrange but it's an option.

 

[mid_gen]

Don't forget you could monetise for people to drop in on your development work for a two hour slot. People will pay to be interactive with someone actually doing AI development to understand more - specifically more senior people that want less BS/slideware and more programming alongside.

You think people will pay to watch me posting on OCUK while Claude agents are building my game in the background?

But I am starting to put together some service offerings, provide consultancy to companies that are unsure how to go about integrating agents into their workflows.

 

[NickK]

You think people will pay to watch me posting on OCUK while Claude agents are building my game in the background?

Not me.. I R Hakzor leet. I'm more interested in some of the cutting edge stuff around optimising AIs - specifically doing things like tensor/manifold compression then breaking it out to allow smaller sub-learning cycles to be applied for addressing security vulnerabilities in the LLM weightings rather than via costly additional feedback loops with security agents sucking tokens to inspect what took tokens to build and then reject it to absorb more tokens to rebuild ... It also saves the planet and reduces the need for data centres capacity..

The key piece is people do the Claude, but don't necessarily feel confident they're doing a right 'pipeline' etc.. There's a lack of maturity in AI as you know, but there's still a lack of maturity in cloud too.. it's a bit of a comedy of errors.. but only hidden by everyone in the same boat.

But I am starting to put together some service offerings, provide consultancy to companies that are unsure how to go about integrating agents into their workflows.

The company ownership side of investment of time and experience helps go far beyond the normal AI consultancy/front deployed engineer. You should get traction pretty quickly.

 

[mid_gen]

Not me.. I R Hakzor leet. I'm more interested in some of the cutting edge stuff around optimising AIs - specifically doing things like tensor/manifold compression then breaking it out to allow smaller sub-learning cycles to be applied for addressing security vulnerabilities in the LLM weightings rather than via costly additional feedback loops with security agents sucking tokens to inspect what took tokens to build and then reject it to absorb more tokens to rebuild ... It also saves the planet and reduces the need for data centres capacity..

I'm not so much interested in the underlying tech, never have been tbh (I'm not a C++ nerd), I just care about the practical applications of it.

The key piece is people do the Claude, but don't necessarily feel confident they're doing a right 'pipeline' etc.. There's a lack of maturity in AI as you know, but there's still a lack of maturity in cloud too.. it's a bit of a comedy of errors.. but only hidden by everyone in the same boat.

I've reached a point now where I have total trust in the Claude Code setup on this project, the architecture is solid, it's proving itself to be robust and scalable as I build out the gameplay functionality. It's basically transparent, I just talk to Claude about what gameplay feature or tweak I want to add and it does it. Quick test...then onto the next feature.

It's no an exaggeration to say this is more than a 10x speed up compared to the pace I was able to develop gameplay at Ubisoft on the gigantic steaming pile known as Anvil/Scimitar.

I'm not actually that concerned about not winning that grant, as I'm now very confident I can get this first game to market by the end of the year. It would be unthinkable without being able to work at this pace, but very achievable now.

The company ownership side of investment of time and experience helps go far beyond the normal AI consultancy/front deployed engineer. You should get traction pretty quickly.

Ultimately, I don't really want to get into the consultancy/services game. It would certainly be a smart move financially, I have acres of relevant, in demand experience and I could be making six figures comfortably, or building a very lucrative consultancy....but I just want to make games, earn enough to be able to retire in 20 years time, and enjoy my life. If I can make enough to build a little studio of 5-6 people and build fun indie games and provide some local jobs then I will be a very happy man.

Any consultancy I do will just be a means to an end to finance getting the studio established.

 

[NickK]

I'm not so much interested in the underlying tech, never have been tbh (I'm not a C++ nerd), I just care about the practical applications of it.

Funny I've gone from deep detail, through to business (not caring about tech), through it more concerned about problem/data aspects rather than the frameworks or technology stacks themselves (ie still not caring about tech... but more the data itself). Tech bores the crap out of me - again like you it's a means to an end (hence I turned off programming in the early 2000s). I just see the understanding of the data being a way to navigate some of the issues. Solves energy, solves waste and thus ends up being cheaper.

If you're building a game company one point could be to make it 'keep what you kill' partnership to share the normal company function expenses between the partners (ie each of those 5/6 being partners). The danger with a traditional hierarchy "employees" in the AI Age is a move away from the traditional to more the entrepreneurial/value creator and for that reward will ultimately be performance linked.