Episode 18
UnityAI and Unreal 5.8: Navigating the Frictionless Future of 'Sentient AI'
UnityAI and Unreal 5.8: Navigating the Frictionless Future of 'Sentient AI'
This episode was recorded in my typical raw, unedited fashion—though I’ve added a video component this time around, mostly because the internet told me to. But let’s be honest: the core of this show is, and always will be, about the unpolished reality of being a solo developer in a world that’s changing faster than we can compile.
Lately, my head has been spinning between two major shifts. On one hand, Unity is making massive waves with their Sentient AI integration, promising a level of autonomy that feels both thrilling and terrifying. On the other, we have the first glimpses of Unreal Engine 5.8 and what it means for those of us already deep in the trenches of the Epic ecosystem. Is the gap closing, or are we just choosing which flavor of AI-assisted madness we want to live in?
We dive into the psychological trap of video podcasting, why I’m resisting the 'high production' allure, and how I’m thinking about the convergence of these tools in my own workflow.
Key Discussion Points:
- The Video Podcasting Trap: Why I’m recording video for discovery but still believe audio is the purest form of this diary.
- Unity Sentient AI: My initial thoughts on Unity's bold move and whether it actually solves the friction for solo creators.
- Unreal Engine 5.8 Roadmap: What I’m looking for in the next major update and how it fits into my current long-term project.
- The Cost of Polish: Why I’m sticking to 'raw and unedited' in an era of hyper-produced content.
- AI as the Bridge: How tools like Whisper and local LLMs are becoming the invisible assistants I never knew I needed.
Tools & Tech Mentioned:
- Engines: Unreal Engine 5.8, Unity.
- AI Tools: Unity Sentient AI, OpenAI Whisper (used for this very transcript), Claude.
- Hardware: My ever-reliable PC and the MacBook M3 Max that still does the heavy lifting for UX.
If you're wrestling with these same choices or just want to tell me I'm delusional for trying to do both, reach out. Let’s keep the conversation going.
Watch the video version of this episode: @cmonk3d on YouTube
EMAIL ME: cmonkxxx@gmail.com
#UnrealEngine #UnityAI #SoloDev #IndieGameDev #AIFuture #DigitalDiary
Transcript
Hey everybody, I'm back and today I'm recording with this additional video component.
Speaker A:I'm recording with my camera as well, but it's gonna be primarily audio, so if you're
Speaker A:just listening to it, you're not really missing much.
Speaker A:But I do want to embrace this, seems like always growing trend of video podcasting,
Speaker A:which I'm really not done with that.
Speaker A:I don't think it's a good idea to mix those two worlds, you know, podcast should be just
Speaker A:audio in the simplest form.
Speaker A:But here we are, I'm recording video because, I don't know, I really don't know.
Speaker A:Oh I know why, I was reading those guides on the internet, those checklists, bullet
Speaker A:points, what you need to do to have your podcast discovered, and you're supposed
Speaker A:to do those video shorts and you're supposed to put them everywhere so people can find
Speaker A:your audio.
Speaker A:But that kind of defeats the purpose, right?
Speaker A:If podcast is audio, where am I supposed to have video from?
Speaker A:I know there are answers to that, but it's, you know, it always, like the solution
Speaker A:to this, it always feels like you need to put some work to it, which is not
Speaker A:what I want to do.
Speaker A:I want to put my work into working with Unreal Engine and things around it.
Speaker A:I don't really want to edit too much video or audio.
Speaker A:I hope you don't mind.
Speaker A:And in fact, I do sincerely believe that this raw and edited flavor of this show
Speaker A:is something that appeals to you because everything now is highly produced.
Speaker A:There are expectations that maybe creators put on themselves that things should be
Speaker A:very polished and there should be high production value.
Speaker A:What I want to do here, and I think I said it before as well, I just want to connect
Speaker A:with you and I want to share with you what I've been up to.
Speaker A:Because if I'm not mistaken, you're also trying to do your best with Unreal
Speaker A:Engine and all those new technologies that we have available.
Speaker A:So that's what we have now, and that's why I'm here for.
Speaker A:Today I want to talk to you about a few topics.
Speaker A:And because I've been away for quite some time, and I'll tell you why, I have a
Speaker A:bit of a backlog, but this backlog kind of collapsed into a very consistent topic.
Speaker A:There are new updates from both Unity and Unreal Engine.
Speaker A:I'm not really that interested in Unity at all, but the reason why I'm mentioning
Speaker A:this is those updates are very consistent.
Speaker A:Kind of.
Speaker A:What am I talking about?
Speaker A:There are updates from Unity that relate purely to AI workflow.
Speaker A:So I guess that's aligned with what I've been talking about on the recent
Speaker A:podcast.
Speaker A:They have now AI embedded into Unity, and I want to talk about that.
Speaker A:And then there's also an update from Unreal, from Epic.
Speaker A:They shared a bunch of features.
Speaker A:I don't know if you saw, there's this new roadmap, and there's a new Unreal
Speaker A:Engine beta or alpha.
Speaker A:You can test it.
Speaker A:And they already shared a list of features.
Speaker A:And I surprise you with just a bit of preparation that I've made.
Speaker A:I printed out a list of features that they seem to be very proud of.
Speaker A:It's not entirely what I was expecting to see, to be honest.
Speaker A:There's one feature that I'm kind of excited about, because it closely
Speaker A:matches directly a need that I have in my project.
Speaker A:But I'll just go through it.
Speaker A:I don't know.
Speaker A:In case you didn't hear about the latest Unreal Engine 5.8, I'm going
Speaker A:to tell you what's there in a nutshell.
Speaker A:We've got smarter world building tools.
Speaker A:We've got, as always, performance and rendering reality checks.
Speaker A:We've got animation and crowd made simple features, so mass crowd and
Speaker A:animation features, performance upgrades, and world building tools.
Speaker A:Pretty good, pretty good.
Speaker A:So the first one, world building tools, we've got mesh terrain.
Speaker A:That's a feature that's really cool for those who are into building
Speaker A:environments.
Speaker A:And a few episodes ago, we've been looking at this report from Epic.
Speaker A:They shared what do people make most money from on the Epic store,
Speaker A:right?
Speaker A:Maybe not most money from, but what is in demand the most, I guess.
Speaker A:And environments are on the top, which was my surprise.
Speaker A:I don't know.
Speaker A:It was to my surprise.
Speaker A:I don't know if you were surprised by that, too.
Speaker A:But it kind of makes sense that people go to Unreal and they
Speaker A:want to create some kind of reality.
Speaker A:And they need to create the whole world, right?
Speaker A:They've got this standard map, empty level, and they just need
Speaker A:to fill it with stuff.
Speaker A:And what do you do?
Speaker A:Well, you can model everything, or you just get your environment.
Speaker A:And people do that.
Speaker A:And I feel like it's obvious first step.
Speaker A:You get your backdrop, and then you build a game on top of that,
Speaker A:or cinematics.
Speaker A:So they have this new updated technology on mesh terrain.
Speaker A:I've seen some demos on YouTube.
Speaker A:They seem to be quite resource efficient, so that's good.
Speaker A:And fun to play with.
Speaker A:It seems like there are now two terrain systems.
Speaker A:So that's going to create a bit of confusion
Speaker A:which one you go with.
Speaker A:I guess the first one, the regular mesh terrain feature
Speaker A:will be replaced by this.
Speaker A:So yeah, we've got PCG workflows.
Speaker A:We've got those vegetation, the new vegetation system,
Speaker A:procedural vegetation system.
Speaker A:It's something I'll be working on for my project within days,
Speaker A:maybe even tonight.
Speaker A:You can create nanite vegetation.
Speaker A:And it's procedural, and you can really
Speaker A:create plants the way you want them to look.
Speaker A:And they look great.
Speaker A:They look absolutely fantastic, in fact.
Speaker A:And in terms of resources, they
Speaker A:don't really cost you too much in your system.
Speaker A:So you can really put a lot of trees, a lot of plants,
Speaker A:and your world will look great,
Speaker A:and your computer will not choke on those leaves.
Speaker A:So that's world building.
Speaker A:Mega lights.
Speaker A:Light is one of those things that, alongside
Speaker A:with sound design, those two things
Speaker A:give so much depth to your experience
Speaker A:you want to create, to the mood you want to create.
Speaker A:And those mega lights, I didn't do anything with them
Speaker A:yet, but it's something worth looking into.
Speaker A:And now they took it out of experimental realm.
Speaker A:So it's becoming production ready, it seems.
Speaker A:So we are playing more with light.
Speaker A:Even if you have not really interesting-looking mesh,
Speaker A:if work that you have on a mesh side
Speaker A:is not that great, if you apply the right lighting,
Speaker A:it's just like in real world, right?
Speaker A:If you wish something to look better,
Speaker A:just apply different light.
Speaker A:And this technology makes things look better,
Speaker A:ultimately.
Speaker A:Now, animation.
Speaker A:I've seen some demos of that new rigging system,
Speaker A:or refined rigging system.
Speaker A:You can do a lot more with rigs.
Speaker A:It's something I know the least about in Unreal Engine,
Speaker A:to be honest, but it's a really incredible portion
Speaker A:of Unreal, where you can do this puppet theater thing
Speaker A:with characters, animals, objects.
Speaker A:And now you can really rig anything a lot easier.
Speaker A:So this idea that, again, you've got environment,
Speaker A:and then you are coming up with some kind of creature,
Speaker A:and you can animate its limbs,
Speaker A:and that is apparently very easy now.
Speaker A:That's something I know I need to explore
Speaker A:for maybe next project.
Speaker A:There is no room in this one for it.
Speaker A:It's very overblown already, but I'm
Speaker A:not going to do any rigging for that one.
Speaker A:But what I'm working on in my project, in fact,
Speaker A:is the mass crowds, is the meta-human crowd system.
Speaker A:And I'm using the old system.
Speaker A:It's hard to believe it's old already,
Speaker A:but it's from the city sample demo.
Speaker A:It's not great.
Speaker A:I mean, it's amazing.
Speaker A:I don't have many NPCs, but probably around 50.
Speaker A:I can stretch it to 100, and it doesn't really
Speaker A:cost me much in terms of resources,
Speaker A:but it's really tedious to work with in some cases.
Speaker A:And I really had to take it from city sample demo,
Speaker A:and exporting of this mass crowd setup
Speaker A:was just really tough, but I did it.
Speaker A:I did it with AI, actually.
Speaker A:I had to surgically extract that mass crowd setup
Speaker A:with all the nuances that were in this demo project,
Speaker A:and it worked.
Speaker A:But I'm really looking forward to this.
Speaker A:Yeah, and then there is some new sculpting features
Speaker A:are also interesting.
Speaker A:I kind of like that, because all those features
Speaker A:that had to do with sculpting and rigging
Speaker A:and the environment, this is truly something
Speaker A:you can play with.
Speaker A:You can just sit down and create something.
Speaker A:You can just model it and make it move.
Speaker A:It kind of feels like fun.
Speaker A:I feel like before we had a lot of features
Speaker A:that you really had to become master at in order
Speaker A:to leverage them, but these new features seem
Speaker A:to be quite playful, and the promise
Speaker A:is that they will be very impressive visually
Speaker A:and in terms of performance, I think.
Speaker A:So that's cool.
Speaker A:Now, I look forward to it.
Speaker A:I will not migrate my project into 5.8 just yet.
Speaker A:It's something I would do a few versions ago,
Speaker A:because I would have been so excited about a new version
Speaker A:of Unreal, and there will be something I'm really into
Speaker A:that I say, yeah, I want this feature.
Speaker A:I'm migrating my project.
Speaker A:But there is a huge trade-off.
Speaker A:A lot of things will break.
Speaker A:Not all plugins will work, and you cannot go back.
Speaker A:So you need to create this copy of your project,
Speaker A:and it will slow you down.
Speaker A:If you're just playing around, if you don't care,
Speaker A:create a copy of your project, continue with 5.8
Speaker A:in this case, and see what happens.
Speaker A:I think I'm going to wait a couple of months at least
Speaker A:before I move, because I do need that mass crowd,
Speaker A:but I'm good where I am right now.
Speaker A:So that's all I want to say about Epic, I guess.
Speaker A:It's not what I expected.
Speaker A:I was hoping that they will release this AI assistant
Speaker A:fully in Unreal Engine.
Speaker A:They didn't.
Speaker A:I guess it is, but still experimental,
Speaker A:and it's really just like a chat board that
Speaker A:tells you how to do things.
Speaker A:But they didn't do anything that actually makes things
Speaker A:for you on your behalf.
Speaker A:There must be a good reason for that.
Speaker A:But I have a workaround, and I talked about it
Speaker A:in the last episode, and I have updates on that.
Speaker A:And I will leave this thread for later.
Speaker A:But I want to talk about Unity, because Unity
Speaker A:got a lot of spotlight.
Speaker A:I thought they stole all the spotlight they could.
Speaker A:They have this AI agent in Unity,
Speaker A:and you can tell it to make a game.
Speaker A:Many people started to be very excited about it.
Speaker A:Many people started to be critical of that,
Speaker A:because, oh, it's going to create a lot of AI slop.
Speaker A:There will be a lot of very crappy games.
Speaker A:And also, it lowers the barrier of entry.
Speaker A:What's going to happen to industry?
Speaker A:I disagree with all of this.
Speaker A:And I'll tell you why.
Speaker A:Well, yes, you can tell your AI agent in your game engine
Speaker A:that, oh, yeah, I want to have this car,
Speaker A:and I want it to move, and I want it to look like this
Speaker A:and like that.
Speaker A:It will do all those things.
Speaker A:But you still need to have a lot of knowledge
Speaker A:and understanding how game structure should be set up as.
Speaker A:And you need to know about the architecture
Speaker A:of this as a software.
Speaker A:You need to have a clear idea what you want to build.
Speaker A:And that unveils this bigger topic
Speaker A:where I can really speak a lot about as a UX designer.
Speaker A:I used to, I still do, I guess.
Speaker A:Professionally, I still do.
Speaker A:I don't work in my occupation right now.
Speaker A:You could be put in front of a task
Speaker A:where you need to design a website.
Speaker A:And you know there's going to be a hero image,
Speaker A:some call to action, some features.
Speaker A:OK, I said website, let's say dashboard, right?
Speaker A:You know there's going to be inside.
Speaker A:You can just easily map that out, no big deal.
Speaker A:It's going to look good.
Speaker A:But after a few hours of using it,
Speaker A:you will see that it's actually horrible.
Speaker A:It doesn't work.
Speaker A:It's uncomfortable.
Speaker A:You're losing a lot of time moving around.
Speaker A:It's not well thought out.
Speaker A:Because, yes, it looks good.
Speaker A:It doesn't work well.
Speaker A:And there's a big difference.
Speaker A:It may be the nicest looking app with nicest looking UI,
Speaker A:but it will be horrible to use.
Speaker A:You'll lose a lot of time.
Speaker A:You'll be frustrated.
Speaker A:And users feel that.
Speaker A:They don't need to be experts in design.
Speaker A:They will feel that design is bad.
Speaker A:And I feel like the same thing applies to game design,
Speaker A:where you get to have this agent that's basically
Speaker A:like a genie in a bottle, right?
Speaker A:You're going to say, oh, make me this game.
Speaker A:I want it to be like this, like that.
Speaker A:And that's it, right?
Speaker A:No, no, it's not.
Speaker A:You need to start from ground up.
Speaker A:You need to first do a lot of planning.
Speaker A:You need to set up all the edge cases.
Speaker A:You need to think about the story,
Speaker A:about the meaning of this.
Speaker A:And when there is this agent asking you, hey,
Speaker A:what do we do next?
Speaker A:What do you want?
Speaker A:You're never going to think about the structure.
Speaker A:You're just going to jump into pipe coding your game.
Speaker A:And it's not going to work well.
Speaker A:It hardly ever does.
Speaker A:It never does, actually.
Speaker A:I'm just going to say it never does.
Speaker A:You can create simple games, and they'll be horrible.
Speaker A:They'll be boring.
Speaker A:They will be not engaging.
Speaker A:Another part of this problem that exists with Unity,
Speaker A:it costs money.
Speaker A:And it costs quite a lot.
Speaker A:Try to make a game.
Speaker A:And if your expectation is you're not
Speaker A:going to spend any money, at some point
Speaker A:you will need to spend some money,
Speaker A:because you will run out of credits.
Speaker A:And you will be facing this question.
Speaker A:Do I want to put more money into this crap?
Speaker A:The answer most likely is no.
Speaker A:I was just playing around.
Speaker A:And then when you need to put money where your mouth is,
Speaker A:where your idea is, you will realize that I don't really
Speaker A:want to spend any money on this, because I'm not convinced.
Speaker A:I'm not committed to it.
Speaker A:And I'm very deep into my own game-like experience
Speaker A:in Unreal Engine.
Speaker A:And I have to tell you, it starts to cost me money.
Speaker A:I'm not paying more than I can afford,
Speaker A:but I'm running out of tokens regularly.
Speaker A:There are days where I need to wait for my AI token
Speaker A:allowance to reset.
Speaker A:And I just need to think about budget a lot.
Speaker A:Now it sounds like I'm relying on AI a lot.
Speaker A:And at this point I do, because I
Speaker A:did all I was able to do on my own.
Speaker A:I've set up my experience, the environment, characters.
Speaker A:I was doing a lot of stuff manually.
Speaker A:But there are things related to code, especially,
Speaker A:that I just need AI to help me.
Speaker A:I could use a person that knows how to do it,
Speaker A:but it would also cost time.
Speaker A:Time is money.
Speaker A:And I don't know such person.
Speaker A:So AI is just a blessing, because I
Speaker A:can move on with this light speed,
Speaker A:working on a project that was barely moving for two years.
Speaker A:So that's basically it.
Speaker A:I'm talking about AI in game making,
Speaker A:because it's my coding partner.
Speaker A:And I recorded this video.
Speaker A:It was a screencast where I was showing my workflow.
Speaker A:And I'm going to take this video down.
Speaker A:I'm going to turn it private.
Speaker A:You can still see it if you follow the link.
Speaker A:But it will not be open publicly,
Speaker A:because I need to record a new version.
Speaker A:I tried to watch this video.
Speaker A:And it was very shallow, very spontaneous.
Speaker A:I don't like it.
Speaker A:It was an hour and a half.
Speaker A:I felt like it was too long.
Speaker A:And I need to make it shorter, more accurate,
Speaker A:and more up to date as well.
Speaker A:Since that recording, I've made a lot of progress
Speaker A:in understanding how things could be better,
Speaker A:how to optimize that workflow.
Speaker A:And I really want to make a new version.
Speaker A:So I'm going to record a new version, maybe
Speaker A:some episodes or parts that you can easily
Speaker A:move around through.
Speaker A:And I think it's going to be a new version to share.
Speaker A:Should I talk about my project?
Speaker A:Maybe I'll just say one thing,
Speaker A:that working on a project like that, or any game, I guess,
Speaker A:the obvious part of working on a project
Speaker A:is the obvious visual thing you can perceive visually.
Speaker A:So how the game looks, how it works.
Speaker A:But then there's the system.
Speaker A:And one of the podcast listeners wrote to me,
Speaker A:and we're in touch, he brought it up.
Speaker A:Building systems is hard.
Speaker A:And it is.
Speaker A:And it's actually the biggest challenge when you make a game.
Speaker A:Because if you don't have it, you
Speaker A:will end up going back and forth.
Speaker A:You will end up guessing.
Speaker A:And the system that's underneath,
Speaker A:just holding the structure of the game,
Speaker A:will put meaning to everything.
Speaker A:So my Unreal Engine game, our experience,
Speaker A:is really advanced right now.
Speaker A:But I don't work within Unreal as much
Speaker A:as I work with the code part right now, with game logic,
Speaker A:with the way to handle dialogue,
Speaker A:with the way how to handle the story.
Speaker A:And while I still consider myself working
Speaker A:on this Unreal Engine project,
Speaker A:I spend most of the time just coding,
Speaker A:and coding with AI, because it's not my field.
Speaker A:Now, as I'm coding with AI, I'm
Speaker A:maybe learning in equal parts as much as I'm coding.
Speaker A:Because the deeper I go, the least forgiving my code
Speaker A:becomes.
Speaker A:If I make a mistake, it breaks things.
Speaker A:It breaks more things that I would like.
Speaker A:So I need to revisit, OK, why did it happen?
Speaker A:How are those things related?
Speaker A:What's the underlying issue?
Speaker A:What's the underlying problem?
Speaker A:And with help of AI, as we investigate,
Speaker A:I'm realizing, OK, so infrastructure of this software
Speaker A:is flawed.
Speaker A:There is a mess under the hood.
Speaker A:There are problems that need to be addressed.
Speaker A:And I have spent so much time fixing the code
Speaker A:and learning how it should have been done
Speaker A:in the first place.
Speaker A:It's an incredible journey.
Speaker A:And I'm both learning and building this thing.
Speaker A:And I just cannot stop.
Speaker A:I cannot stop.
Speaker A:And lately, I'm sitting in front of a computer so much.
Speaker A:And I'm creating so much in such a weird flow.
Speaker A:I mean this flow that I didn't feel for a decade, at least.
Speaker A:And I had no idea that it would ever come back to me.
Speaker A:Because it feels like I'm a teenager.
Speaker A:And I just discovered a hobby.
Speaker A:I feel like I have this clean slate of creativity
Speaker A:that I never experienced.
Speaker A:Because when you use AI tools and if you use them
Speaker A:in the right way, not to create slope
Speaker A:or not to just write code, you see that every challenge can
Speaker A:be solved very quickly.
Speaker A:And you end up moving faster.
Speaker A:You end up just deploying a lot more and fixing problems
Speaker A:instantly.
Speaker A:You don't need to spend time on researching a solution.
Speaker A:You don't need to watch endless videos
Speaker A:to find solutions to problems you have.
Speaker A:You don't need to guess so much.
Speaker A:And that accelerates progress to the point
Speaker A:where I don't need to work so much,
Speaker A:but I cannot stop because I'm seeing the progress is
Speaker A:exponential almost.
Speaker A:And the deeper you go, the more you learn.
Speaker A:And it's just fascinating.
Speaker A:I might be a little bit vague here,
Speaker A:but I also don't want to go into this rabbit hole
Speaker A:where I'm sharing more information that you require.
Speaker A:But if you're interested in any of this,
Speaker A:if you want me to break down any of those topics,
Speaker A:if you would want me to make another tutorial about AI
Speaker A:powered workflow for game design, let me know.
Speaker A:And I'm happy to share.
Speaker A:I think I'm going to stop this episode because there's
Speaker A:more and more people coming into my vicinity.
Speaker A:And I get a little bit distracted.
Speaker A:It's also getting a little bit dark and cold.
Speaker A:So I'm going to end here.
Speaker A:And I'm going to get back to work, honestly.
Speaker A:I'm so glad this is just one take.
Speaker A:I don't need to edit this.
Speaker A:And I'll talk to you soon, I promise.
Speaker A:I will not let you wait as long as it
Speaker A:has been this time around.
Speaker A:I hope you're doing well.
Speaker A:Let me know in comments.
Speaker A:In this video, there is a link.
Speaker A:Go to the video.
Speaker A:Leave your comments.
Speaker A:Share your progress.
Speaker A:Tell me what you're working on,
Speaker A:how you found this podcast.
Speaker A:And I want to reiterate that I make it
Speaker A:because I want to connect with people.
Speaker A:I want to exchange information,
Speaker A:knowledge, experience, inspiration.
Speaker A:As much as I lean on AI, I don't want this
Speaker A:to be my only coding partner.
Speaker A:I want to connect with people like you.
Speaker A:And I want to feel like there is some kind
Speaker A:of a feedback loop.
Speaker A:And it's just fun to talk to people.
Speaker A:People are reaching out to me.
Speaker A:There has been a handful of people already.
Speaker A:And we are in semi-frequent communication.
Speaker A:It works.
Speaker A:And if this continues, I will probably set up
Speaker A:some kind of a group chat.
Speaker A:And I will be engaging in it because I love it.
Speaker A:I do.
Speaker A:So right now, I'll be ending this episode.
Speaker A:I'll be going back home.
Speaker A:I will try to do some more work.
Speaker A:This evening, I will actually work on my level,
Speaker A:on the environment.
Speaker A:I need to put some mega plans.
Speaker A:In fact, I'll try to record this.
Speaker A:And I'll see.
Speaker A:Maybe this video will be released sometime soon as well.
Speaker A:But yeah, leave the comments.
Speaker A:Subscribe to our podcast.
Speaker A:Subscribe to my YouTube channel.
Speaker A:And let's see where we are in a week from now, shall we?
Speaker A:Thank you for your kind attention.
Speaker A:Take care.
