Dr Dharam Shadija on AI, GTA VI, PS5 Pro and the Future of PS6

Dr Dharam Shadija is Head of School of Computing and Digital Technologies at Sheffield Hallam University. He oversees the world’s largest test lab for PlayStation 4 and PlayStation 5 models.

In an exclusive interview with GoHumanize.ai, Dr Shadija explained why artificial intelligence, from Generative AI through to machine learning techniques, will be a core part of the design for the PlayStation 6 in a world where adding raw power is becoming more and more expensive.

Speaking ahead of the release of Grand Theft Auto VI, he also gave his view on whether the PS5 will cope with the demands of the title, and why the PC version, set to release in early 2027, could bring about a suite of enhancements only available to the AI-boosted PS5 Pro.

“I lead a large computing department at Sheffield Hallam University,” Dr Shadija told GoHumanize.ai. “We’ve become known for our work on video games, and we’ve heavily invested in PlayStation as a department.

“We have the largest PlayStation 4 development lab in the world and one of the largest PlayStation 5 labs in the world.

“The kits themselves aren’t all that exciting to look at but we do get our hands on the consoles before they land on shelves, and we’re lucky enough to have a close relationship with Sony, and we do a lot of game development ourselves.

“That's primarily through our students. and we've got an in-house game studio called Steel Minions and that has the publishing rights to publish to the Sony PlayStore. We've got the commercial licenses and everything else, so we delivered the world’s first student-made PlayStation 4 game and the first student-made PlayStation VR game to be sold through the Sony PlayStore.”

Read Dr Dharam Shadija’s full interview with GoHumanize.ai below.

Do you think we’ve already reached peak console?

Have we reached peak console? For more casual gamers, we will never again see the level of change that came before with the visual leaps from PS One to PS2.

Even if you look at PS3 to PS4, it was still quite obvious to most gamers the shift, whereas from PS4 to PS5, most players can hardly tell the difference, and so what we’re now seeing is a shift away from raw graphical breakthroughs and into improvements in frame rate, responsiveness, light quality, complexity of simulations, and into live streaming and more connectivity.

For most players, I think we potentially have already reached a peak, and they won’t see too much change from the PS5 to the PS6, not compared to past expectations for the difference in console generations.

If you look at the difference between the PS5 and the PS5 Pro, it is more around how AI is used to provide an enhancement rather than a significant change in the actual hardware, and after all, the hardware is only becoming more and more expensive, which is an interesting situation too.

For the first 20 or 30 years of console gaming, we saw the hardware costs reducing. Now we’re getting to a point where the hardware costs are actually growing;, that’s partly because of how the world is scaling up AI and the increased demand for components.

RAM, for example is quite high in costs now, and you can’t really go and price a gaming console at $1,800. Even a console in the range of £$700-800 is possibly the absolute maximum, and even then you start to lose that mass market appeal, and that’s a category of users the industry could lose as a result of more expensive consoles.

You look at how much money and resources areresource is being invested in a game like Grand Theft Auto VI, and developers and publishers need a large player base for that investment to pay off, so a world where gamegames consoles become too expensive for most players to buy could destroy that model.

Rockstar have spent a lot of money on creating GTAVI, and they need and expect the returns to come through consoles, or at least a significant proportion of the sales they need, and we’re talking about a video game industry that’s now bigger than films or music. It’s interesting where we are heading if consoles are no longer this cost-effectivecost effective way into gaming.

For someone who doesn’t have a console to play GTAVI, should they invest in a PS5 Pro or will the standard PS5 model still be able to cope with such a demanding new release?

Fundamentally, it comes down to the user experience you want. When you're investing in gaming hardware, that's the key to your decision making.

The standard PS5 will work well enough for GTA VI but there are improvements you get with the PS5 Pro that will give players a better, higher-end and different experience because of the frames per second and so on.

That’s what you need to consider when you’re asking the question as to whether you should invest in a PS5 Pro as opposed to the PS5.

If this was five years ago, I would have told people to wait because prices will come down, but unfortunately that’s not happening, and that’s driven by demand on components from elsewhere, outside of gaming.

The good news is that the PS5 has a long shelf life ahead of it. It will be a good enough console to sit under TVs in homes around the world and it will remain a primary platform for a long, long time, probably until at least early 2030. I expect Sony will continue to support the platform until at least 2030 but will it give players the user experience they are expecting from the trailers for games like Grand Theft Auto VI?

If you are looking for a more responsive, richer experience, with higher frames per second then the PS5 Pro will do a better job, and if we look at the life cycle of Grand Theft Auto V, we may well see enhancements in the future that only the PS5 Pro can handle.

We know that games are no longer released these days in one fell swoop and that’s it. GTA VI will almost become a platform in its own right with additional releases, modes, content and other enhancements, and the PS5 Pro will enable you to actually run everything much more easily.

The PS5 will remain a good option for players, at least for another three years or so.

Given the likely constraints on hardware due to costs, could we end up with a PS5.5 instead of a true PS6 once the next generation comes round, with AI technology doing the heavy lifting?

I think the short answer is yes, and I think the reason for that lies in the PS5 Pro and its use of AI rendering and other features. A lot of this is patented stuff. When our students need to access the SDK they need to sign an NDA, and Sony are understandably very protective of this stuff and how much information gets out about it.

For PS6, there will be logistical challenges around the cost of memory and the need to keep it at a price point where it will still have a big market, even as a higher-end console experience. Unfortunately the economics go against it being something more.

This is why the PS5 will have such a longer shelf life than anyone expected but we’re all excited and we’re hoping to get our hands on dev kits for the PS6 in 2028, which will then allow us to set out a ladder for commercial launches. That’s based on when we received the PS4 and PS5 in the past.

I suspect they’ve launched the PS5 Pro with a perspective to try and extend the shelf life of the PS5 era even further. Sony and Microsoft still make a lot of money from console sales, and they have invested heavily in the next generation to come, but it may be inevitable that what we get feels more like a PS5.5, but that may be an opportunity.

Backwards compatibility, being able to play games across the PS One, PS2, PS3, PS4 and PS5 on the newer platform will be quite important to making the PS6 feel like the next step forward.

We think, and I suspect, that they're going to try and remain true to the high-end console environment with the PS6, no matter its specifications, but that may come down to better optimisation, and on-board AI features, that enable Sony to push the hardware further.

I think even now with the PS5 and especially the PS5 Pro, the key for developers is to make the most of the current hardware rather than waiting and hoping for more processing power in the future or more RAM, because actually there’s more to find in the existing hardware.

If I was a betting man, I think the PS6 will build on what the PS5 Pro is already doing but we know the market will respond as well to a console bearing the name of the PS6 more so than just the PS5 Pro. It doesn’t make that same big headline story as a brand new generation for PlayStation.

I suspect the reactions to the differences in experience of playing Grand Theft Auto VI on the PS5 compared to the PS5 Pro will be useful for Sony to judge.

Will the next generation make it easier and quicker to make games or will the greater complexity lead to players waiting 15 years or more for the next GTA?

In the case of the Grand Theft Auto series, that’s not down to the complexity of the consoles and hardware. Rockstar just wanted to take their team because they have the money to work at their own pace, but will the next advances in hardware lead to longer waits for sequels from other developers?

The short answer is yes because it will become even more important to optimise games to use the hardware to get the best, most effective performance, making the most of the right capabilities rather than just relying on more power, and that can become more complex.

We saw this as the industry moved from the PS2 to the PS3 and then to the PS4 and PS5 but this will be compounded for the PS6 in terms of understanding the hardware challenges, such as minimising heat, especially in a home environment.

It’s something we have had to address in our labs. We’ve got about 30 different kids in a single large computer room and since the PS4 we have had to account for heat emissions from these machines, and especially with the PS5 lab. This is the kind of complexity I’m talking about these days.

Developers are having to spend longer and longer tackling these problems but it depends which end of the market you’re trying to cater to because we release games ourselves, and these student-made games run perfectly fine. They’ve been through Sony’s QA process before they are sold on the Sony Play Store. So you can produce games quickly for these platforms but the high-end gaming experiences, with all the bells and whistles, are still made to run at the highest frames per second possible, it is only becoming a much more rigorous process.

The other side of this is the amount of content in these games too. You compare them to a movie, even a large-scale Hollywood blockbuster that takes a couple of years to make, and games are taking far longer because they simply contain more storytelling, more levels, more characters, more writing, more dialogue, more of just about everything. That has only increased since the PS2 era where people expected games to be created and released more quickly than they are today, such as with the older Grand Theft Auto games.

Over the last 10 to 15 years there has been far greater collaboration between different specialist teams to add more depth to the games and that increases the amount of work required too. So even before considering the hardware complexities, it takes longer because more is going into games. If a movie takes two or three years to be created by a studio, think how many movies worth of ideas and missions and stories go into a game like the new Grand Theft Auto games. But you’ve also got to make use of the hardware to get the best performance possible, and bring all these ideas to life, and test them, and suddenly it’s not so surprising these games are taking longer and longer to make.

What’s on your wish list for what we need to see from the PS6 to make the next generation feel worth it to players?

So there’s a few things I think we need to see in terms of a PS6 wish list.

First of all, larger unified memory… but still keeping the costs at the right point. I think 24GB to 32GB would be up there.

A powerful, dedicated AI processor is now a must have feature rather than something that would be nice to have.

Advanced hardware for ray tracing will be quite important, as will be hardware-supported frame generation.

You look at the size of Grand Theft Auto VI, which is going to be about 200GB, and the next generation of consoles are going to need a much larger SSD, so 2TBs or more.

Full backwards compatibility from the PS One through to the PS5 has to be the promise of the PS6 to consumers.

Lower power consumption will be important too and last but not least, and this very important for the home environment, is making sure the PS6 is a quieter console too, and part of the problem they’ll have to solve to achieve that is heat generation.

We all know what it’s like working on a laptop or playing on a console when suddenly the fan kicks in. You don't want your console to sound like a vacuum cleaner at home. So the PS6 generation of consoles must be a step forward in low heat emissions and heat management and I suspect Sony have come up with some clever ways to handle that.

However, at the end of the day, everything comes down to trade-offs, and what Sony is willing to compromise on from the perspective of providing the best user experience possible.

Will the average player care so much about frames per second or ray tracking and so on? It’s about how Sony take whatever wish list they started with and decides what they can deliver at the right price point to keep the idea of a console viable.

When we compare what is now possible with a high-end gaming PCs when playing games like Cyberpunk 2077 and other visually intense titles, will players have to temper their expectations for the PS6?

I was watching the Nvidia launch and it’s impressive what even these super powerful, dedicated gaming laptops can do but they are coming on to the market at £2,500 or $3,200 or more, and they really are at the higher-end of the market.

However, if you step away from the raw technical details and consider a company like Sony has the purchasing power to go and negotiate prices for their next mass produced console product, and that’s where economies of scale can make a real difference. There is a balance they can strike there.

I think it will come down to how well the PS6 is able to cope with AI upscaling and other new features. That is going to be key. AI-assisted rendering, latency reduction, a seamless lack of loading screens, improved streaming, better ray tracing and lighting, and all these other areas of focus where we expect the PS6 to make its mark rather than trying to compete with high-end gaming PCs.

These top-of-the-line gaming laptops and PCs will always have an advantage in terms of their technical specifications but there is a cost to that, and that is not the point of consoles.

Games consoles tend to be a birthday present or a Christmas gift and even now at £600 or $800 that is a lot of money for a Christmas present. Spending even more than that on a gaming machine is a very exclusive experience that is only open to a small portion of the market.

With a console, you optimise a standardised bit of kit made up of hardware that may not be as impressive on paper, but you may be able to get more out of it, and it’s quite interesting where that could go next with AI, especially machine learning techniques. There are performance advantages available there, and that is quite exciting.

That’s why, for example, Sony have invested so heavily into AI with the PS5 Pro and their own patented tech which has enabled them to find performance gains to take games from 30 frames per second to 60. I expect that we’ll see an update release for Grand Theft Auto VI in early 2027 to take advantage of these features once the PC version is released. That might be a moment that pushes more people to buy the PS5 Pro who already own the PS5, even before the release of the PS6, because of the greater experience.

I have a feeling we’ll get an upgrade to Grand Theft Auto VI for the PS5 Pro, much like the first-person mode that was only available on the PS4 after it was originally released for PS3. That would be a very compelling reason to invest in a PS5 Pro.

What are the sacrifices you expect Sony to make in order to deliver on the promise of the PS6?

I think for a device that's going to be used in the home, managing heat and noise is really important, so the memory bandwidth will be a key challenge. You can’t necessarily just put more into the PS6 for obvious reasons around cost too so they will have to be very intelligent about increasing the usable memory, and that’s where machine learning techniques could come in.

We’ve reached the point of diminishing returns when it comes to Moore’s law where you can’t just put more and more power into a games console and actually get better performance, because of the heat produced.

Our labs are fully air conditioned to try and keep the heat away and that’s never going to be a solution in a home setting. People don’t want a mini data centre under their TV giving off heat and noise emissions.

We’ve been fairly innovative as an industry, always trying to move to the next bit, and looking for solutions to that, and we are getting to a point where things are smaller and smaller, and even more powerful, but now the exciting new frontier in this space is AI.

There will be other, new hardware breakthroughs along the way but it’s also a question of how much gaming can afford to tap into these advances versus the wider economic forces demanding these resources too.

Could Sony follow Apple in creating their own silicon designed specifically for PS6 for the potential performance gains and lower power consumption?

Every company will invest in their own research and development but I suspect Sony won’t commit to following Apple’s path so directly. Sony aren’t going to necessarily make their own silicon now because that’s what Apple did but there are lessons they can take from that success.

AI specific hardware for the PS5 Pro and the PS6 are two areas that Sony have already invested into and that will remain key. You look at a MacBook today with Apple silicon and it out-performs the older MacBook Pro models. That will be the ambition for the PS6, so let’s hope it comes sooner rather than later.

Should there be any concerns about PS5 models continuing to work long after this generation was originally assumed to have ended?

We've not seen anything to worry about in that sense across the 30+ kits in our lab. We’ve not seen any changes in performance, and as the person who signs the cheques I have not had to make any further requests for additional machines.

They’re very well built, as is the PS4, which we’ve had for quite a while now. These are dev kits but in terms of processing power, it’s the same as the console you can buy as a consumer. They just tend to have a bit more memory compared to the machines that gamers use but we use these machines from 9am to 5pm every day of the week and we don’t see them breaking down.

Nothing's come through to my desk yet in terms of any need to replace this or that. I must say we’ve had to do more work to keep our Windows PCs going than the PlayStation kits all sat next to each other, for either the PS4s or PS5s.

What are the biggest misconceptions and urban myths around dev kits?

I think the biggest myth is that these are some sort of super consoles and that isn’t the case.

What we do get with the dev kits is a little bit more support in terms of debugging, profiling, memory analysis and performance tracking. That’s all very important when you're developing these games. It doesn’t have a purpose for playing games.

Our kits do have larger memory than the consoles at home because it needs to run all these additional diagnostic tools in parallel to see how the game is working and because of that they do cost a lot more, around £2,000 each for a PS5 dev kit.

The other big myth is about how they look because we are not allowed to take pictures but they don’t look like a shelf-ready product.

It’s always exciting to get our hands on the next generation of dev kits rather than relying on some form of emulation because it is vital to learn the hardware and how to get the most out of it, especially for our students. To get these developers of the future working on the real hardware, that makes a real difference.

Are you worried for what the cost of the PS6 dev kits could turn out to be?

I expect it will be a magnitude more expensive than what we paid for the PS5 dev kits. They could cost a grand or two more on top of what we currently pay because of the additional bits, particularly given the cost of memory is going to play a big part in that.

For us, it’s an essential spend. We’re in education so for us it’s about ensuring we are producing the best possible students who are employable or can go and establish their own companies, and that requires that we give them full access to the right kinds of kit. That’s important to us but yes it is going to be expensive.

We have already started to work on the business case so that we can invest in the new equipment in a couple of years. It was the same with the PS5, although the dev kets were delayed due to COVID and the chip shortages.

They weren't cheap last time and I don't think they'll be cheap next time either but they're so essential in terms of the world of work from a student's perspective. Emulators and other forms of simulation just don’t cut it.

If the rumours of a handheld element to the PS6 turn out to be true, in order to compete with the Nintendo Switch 2, how much more expensive and complicated would that make the dev kits?

If the PS6 is a handheld or partly a handheld, that would require a different way of doing things.

In terms of the dev kit for a console like that, given what they are expecting game developers to do, it would have to be compatible between all the hardware elements, and that will be an issue.

That would impose a limit on what you are able to do, and the amount of time that you can spend developing and redeveloping a game for both a normal console and a handheld if that’s the way they decide to go.

There would be a lot more work involved, and I suspect development timescales for a console like that would be equivalent to making the same game to launch on two very different consoles, and take a lot longer as a result.

The reason for this is the hardware limitations which will be more to do with the cost of a handheld rather than trying to deliver the same performance between both devices. This is why I suspect they won’t go down the handheld route. That would almost feel like a step back.

The Nintendo Switch and Switch 2 are a different proposition to the PS5 in terms of performance, let alone whatever the PS6 and the Xbox equivalent will be, so if I was a betting man I don’t think Sony would go that way, and I hope they don’t. The cost to set up a PS6 lab is going to be high enough without having to budget for a range of different dev kits across multiple pieces of hardware.

Does true backwards compatibility become a non-negotiable for the PS6?

Think in one sense, yes because there are generations of great games that can become a library for the new consoles immediately, and a lot of gamers have spent a lot of time with older games and remain devoted to these titles, so backwards compatibility has to be a key thing. Backwards compatibility for PS3 titles in particular could be an interesting selling point for the PS6.

PS3 integration has remained a challenge, even for PS5. It could be key for the PS6 if it can achieve compatibility, accuracy, stability and certification across thousands of titles, but that will remain a challenge.

Could removing disc drives from all PS6 models and committing to digital distribution give Sony some extra space inside the unit they could use to boost performance?

It’s an interesting idea because we’ve already got to a point where we might assume everything will be accessed through the cloud once the PS6 is released.

Discs are already being treated as a thing of the past so actually having any sort of physical drive could be replaced in favour of other hardware in the design, even if it adds a bit of cost, or perhaps they could use that space for something else, like cooling that sort of device.

It could also allow Sony to reduce the size of the PS6 and save space because houses aren’t getting any bigger, and there isn’t always a lot of space underneath your TV. You can’t keep going bigger and bigger just for the sake of performance. It’s about how you can use that space more intelligently.