As a long-standing mentor across several programs, Chloé is passionate about supporting the next generation of design talent and helping more women move into leadership roles. As part of our 5 Questions series, I caught up with Chloé to discuss career imposter syndrome, pet peeves and the 24/7 drive of the creative mind.

As you'd expect, the show itself was sprinkled with spectacle, pizzazz, and plenty of Disney magic. Bart Simpson opening proceedings. Stitch arriving on the big screen with his new pink co-star. Giant red balloons bouncing around the arena for a mass game of Keepy Uppy. Fantastic.

But D23 isn't just about feelgood vibes, it's mainly about announcing new projects. And the list of forthcoming films made me decidedly underwhelmed.

What's coming next

In the animation corner: Frozen 3. Incredibles 3. Coco 2. Zootopia 3. A sixth (yes, sixth!) big-screen outing for Ice Age. A second Simpsons movie. In the live action corner, meanwhile, we're getting a remake of Tangled, and a sequel to 2025's previous live-action remake Lilo & Stitch.

If you count up the films Disney and Pixar announced, only four are original: Pixar's Ghost Market and Gatto, and Disney Animation's Clay and Hexed. The rest is a company repurposing its past.

I've written before about why I think Disney's live action remakes are all wrong, but a live-action Tangled is really depressing. The 2010 original was already made with CG, so turning it into "live action" – which will doubtless be dominated by CG anyway – feels less a brave reinvention and more like a tweak of the 3D software.

Not to mention that when it arrives in March 2028, it'll be remaking a film that's barely 18 years old. One that very much hasn't aged, and anyone who's interested will have seen on Disney+ already.

What's depressing me most is that I can see the business reasons for what they're doing. The live-action Lilo & Stitch took over $1 billion at the box office in 2025, so a sequel is a slam dunk.

Original animated films are expensive and risky, as Pixar learned when Elio struggled at the box office last year. If I was a Disney investor, I'd be screaming "Make Frozen 3!" from the rooftops.

But over the long term, that just means Disney will end up eating itself. If you think about it, every franchise Disney is milking today was once a risky original. Where are the next risky originals coming from?

The good news

Admittedly, there is some new stuff on the way. Hexed, Disney Animation’s next film, is an original that hits cinemas in November. At D23, Disney announced two big additions to its voice cast: Jodie Foster as Queen Celeste and Walton Goggins as a character called Beef Roger Crummchuk. That name alone makes me want to buy a ticket .

Pixar, meanwhile, has Gatto, a Venice-set adventure from Enrico Casarosa, the director of Luca. The latter's sunny take on the Italian Riviera was charming; a shame it went straight to Disney+ in 2021. At least Casarosa is getting a proper cinema release this time, with Gatto due in March 2027.

Further out, there’s Ghost Market, a totally original Pixar film due in March 2028, and Disney Animation’s Clay, also set for 2028. Disney's said very little about either, and Clay is only a working title. But I'm still eager to learn more.

Finally, there's The Bluey Movie, and while that's based on a well-known character, a jump from TV to the big screen can often lead to great things: just look at Aardman. The irony? Disney didn’t make it: they're only distributing it.

Yes, one of the most exciting projects at Disney’s event was an idea someone else had, and which somebody else is making. In the meantime, I'm putting my trust in Laika to bring some child-friendly movie magic back to our cinemas.

Almost three months on, and with the new packaging designs out in the wild, a more practical concern is causing problems for some. The standard Coke looks more Coke, but so does Coke Zero, making it potentially easier to confuse the two. If one of the aims of design is to improve the user experience, some see that as a fail. Or is it part of a longer-term plan?

Coca-Cola Zero was launched back in 2005 and rebranded as Coca-Cola Zero Sugar in 2016. The packaging was initially black, and then it became red with a black logo. Now only a black ribbon remains as a colour clue from the visual system that distinguished the product from standard Coke.

That might seem like a minor inconvenience. The packaging says 'Zero sugar' in fairly large type, so it shouldn't take too long to find the product you want. But it seems it can be a problem for people who have to select the product quickly and repeatedly as part of their job. An Australian flight attendant recently posted on TikTok to complain that the change had made her work more difficult.

In a post showing two almost identical-looking cans, she writes that she's “lost count of how many wrong Cokes l’ve handed out this week”.

The post received thousands of likes and over 100 comments. Some of those note that the confusion can have potentially serious consequences for people who have allergies or certain medical conditions.

“I’m allergic to an ingredient in Coke Zero but can have regular coke. Why would they think this is a good idea?!” one person asks. “I RELY on that black lettering for Zero cuz I’m diabetic. Why on earth would they do that?," another person writes.

“I can’t tell yall how many times in a rush I’ve accidentally grabbed Coke Zero and didn’t mean to,” is another comment.

According to Coca-Cola's press release at the time of the rebrand, the main objective was consistency. “With one seamless visual identity, we can deliver one recognizable brand experience to consumers,” it said.

It seems that the brand experience some customers would most like to have is to be able to select the product they want without having to hesitate.

There is a logic to the evolution of Coca-Cola Zero's branding. When it was first introduced as the biggest new Coca-Cola product launch for two decades, it needed the all-black can to communicate its existence.

But people were confused about what Zero was, and about how it differed from Diet Coke. That led to the last design refresh in 2016. It was clarified that the product was 'Zero sugar' rather than Zero cola or zero anything else, but it was also decided that the distinction of the black can was no longer needed.

Now people know that Zero Sugar exists and generally know what it is, perhaps it doesn't need to look so distinct. But why make things more difficult for the consumer?

Some think the explanation goes beyond the question of consistency. Over on Reddit, there are suggestions that the decision to blur the visual distinction between the two products is part of a long-term plan to ease customers off the standard sugary coke, and possibly even discontinue it to make Zero Sugar the standard product.

With sugary drinks taxes and labelling rules coming into force in various countries, it could make sense to try to push people towards the zero version. Nevertheless, Coca-Cola has never expressed any intention to phase out sugared Coke, and it's recently started making a sugar cane version in the US. Most standard Coke in the US is still made with high fructose corn syrup, while in Mexico and Europe it's made with sugar cane or beet sugar.

To remedy this, Bentley devised a synthetic soundtrack built from drum recordings modelled on its old eight-cylinder engines. Distinctly different from its previous models, the startup noise replaces the engine purr so many drivers are used to, but will this mark the death of the traditional driving experience, and is it enough to convert classic car lovers?

To get a deeper understanding of this new EV trend, I caught up with Max De Lucia, co-founder of sonic branding agency DLMDD, to discuss what the future of EV sound design will bring.

There was the M4 iPad Pro launch video, in which Apple crushed a haul of cameras, musical instruments and other creative tools to make the point that the tablet is very thin. Then there was the Thailand advert, packed with stereotypes, and the cringey Parent Presentation, a campaign that aimed to teach students how to convince their parents to buy them a Mac.

More recent ads suggest that Apple's learned from those blunders, and it's avoiding the mistakes that other companies are making as AI slop takes over social media. An embroidered Apple logo invite for its upcoming Welcome Home event is the clearest sign yet that it's pulling out all the stops to ensure that it's associated with hand-crafted human-made creativity.

Apple's Welcome Home event scheduled for Tuesday (13 October) has come as a surprise to some since the company hasn't run an October event for the past couple of years. That said, we had been expecting news about new home devices. With September's event already so packed with the iPhone Duo reveal alongside the iPhone 18 Pro and Apple Watch Series 12, it made sense to put those back.

The advert for the event posted by SVP Marketing Greg Joswiak above shows the creation of the Apple event invite. This almost always involves a depiction of the Apple logo in some form, depicted in a way that alludes to a theme or piece of news that will be revealed at the event.

The Welcome Home invite takes a distinctly handmade approach. It sees the iconic 1977 rainbow Apple logo revived in an embroidered design. The stitching is the work of Sydney-based twin sisters and multidisciplinary artists Maricor/Maricar, who bridge the disciplines of design, illustration and embroidery (you can see more of the process in their Instagram post below).

Embroidery is an art form that dates back thousands of years, yet Apple using it for an advert in 2026 is breaking news. It's the latest example of the lengths the brand's going to to make to make it as visible as possible that its marketing is made by humans, not AI.

From Skechers sketchy AI posters to the frightful Fanta Halloween advert. some brands seem to have underestimated the strength of feeling against AI. Despite being part of the AI industry (albeit a relative latecomer when it comes to developing its own models), Apple's seen that. It's recognised the growing backlash and the desire to connect with more traditional human qualities. It's now creating a brand around human craft.

Other recent examples have included the stop-motion animations created for the launch of the new Apple mini and Apple Games (see below)

Apple's also realised that behind-the-scenes videos are now as important as finished adverts in the AI era: or perhaps even that behind-the-scenes videos are now the real advert. If your brand is going to invest in human-created art, you want to make sure people recognise it as such and appreciate the craftsmanship that went into it.

Meanwhile, iPhone Duo pre-orders begin on Friday 16 October. Alas, there's no news that preorders will be available for the embroidery.

While some predicted that AI would be the final death for handwriting, a strong counter movement has been pushing back against that idea. People are increasingly rediscovering the benefits of handwriting for focus, reflection and creativity. Businesses have even started hand-writing social media posts amid the backlash against AI slop flyers on Instagram.

And right now, you can get a lovely pen for less then the price of a month's subscription to ChatGPT. Amazon has various versions of the legendary Parker Jotter (as used by James Bond in Goldeneye), on deal. The classic black and steel model with arrowhead clip is reduced by 30% to $8.34 and the larger stainless steel Jotter XL by 18% to $32.79.

If you prefer to go even more traditional, the Parker 51 fountain pen favoured by the late Queen Elizabeth (and used by James Bond in the Moonraker novel) is reduced by almost $25 to $101.25.

Here are all of the best deals.

• Shop Parker pen at Amazon

Not in the US or want something different? See more pen deals below:

In a subtle yet touching homage, shoe brand New Balance has recreated one of Jobs' most iconic looks, honouring his legacy as a creative inspiration. Quiet and authentic, New Balance proves that honouring the past can be done with small yet thoughtful actions.

Appearing among promo shots for the grey 992s, New Balance subtly recreated Steve Jobs' iconic pose from the original iPad presentation in 2010. A key part of Jobs' 'uniform', the quiet homage is a tasteful Easter egg that honours the icon without grand gestures and loud design. Minimalist, just as Steve Jobs would want it.

"Proof that the best branding doesn’t need to shout. If you know, you know. Brilliant little detail," one X user praised. "The kind of details I adore," another added.

For more creative inspiration, take a look at how Apple became the world's most valuable brand or check out the best Apple ads of the last 50 years.

A classic graphic design reference book has resurfaced on social media, and it's getting attention for the familiarity of some of the symbols it depicts. Some of them look incredibly similar to identities belonging to modern-day companies, from AMD to Google.

The book is Trademarks & Symbols, Vol. 2: Symbolical Designs by the Japanese designer Yasaburo Kuwayama. While Volume 1 had focused on alphabetical and letter-form designs, the second tome centres on symbolical and pictorial designs. It's a directory of real symbols and logos organised according to their form, for example circles, squares, trees, eyes and human figures.

From a just a few pages shared on social media, people have spotted designs that resemble the logos of chip maker AMD, Google Drive, Chase Bank, Natwest, Canadian Imperial Bank of Commerce, Portugal's state bank, Renault and more.

Have the graphic designers behind modern big brand logos been picking and choosing old designs to recycle? Probably not. Geometric forms are relatively universal, and Yasaburo's book is so extensive that it's logical that modern brands could arrive at similar shapes on their own.

One observer has pointed out that in the book's appendix some designs are listed as being used by multiple companies. Even today, we see brands with similar logos.

When it comes to minimalist geometric symmetry, there are only so many options. Nevertheless, if you're looking for a logo design for your startup, some people reckon the book still has a few gems that could be available for the taking.

Trade Marks & Symbols was originally published in Tokyo by Kashiwa Shobo and Orion Press and internationally by Van Nostrand Reinhold Company. You might have to part with a lot of money to get a copy today. There's a hardcover copy selling on Amazon for over $2,000.

The nearest contemporary equivalent I can think of are Jens Müller's Taschen books Logo Beginnings and Logo Modernism, which I can highly recommend.

I really did think the iPhone Air was for me when Apple announced it last year. As an iPhone mini apologist, I've always favoured design over features, whilst lamenting the trend towards massive, heavy bricks. But all of those compromises that tech nerds have accused the Air of making in its pursuit of ultimate thinness? Yeah, they're valid.

I still love the design of the iPhone Air. From the ultra thin profile to the polished titanium edges, this is the most premium looking and feeling iPhone ever. It's the closest thing we've had to Steve Jobs' dream of a device that's genuinely all-display – the phone is so thin and light that it disappears into the background when looking at the 3,000 nit display.

But yes, it only has one camera lens. And despite Apple's claims that it's somehow worth two (still not sure how that checks out), the limitations are a pain in everyday use. Countless times I've tried to take a 0.5x photo, only to remember the feature doesn't exist on the Air. Same goes for zooming in on anything further than the near distance – the digital zoom just doesn't cut it.

The other major drawback, of course, is battery. Over the course of a year, my iPhone Air's battery health has degraded to 91%, which is much quicker than my iPhone 15 Pro degraded. The result is not-quite-all-day battery life, which has got me into trouble a few times when travelling.

And while I said I love the design of the iPhone Air, I do take issue with one aspect: the width. Why, oh why, is the so-called thinnest, lightest and therefore most portable iPhone too wide to use one-handed? It makes no sense to me that the iPhone Air is taller and wider than the iPhone 18 Pro.

Speaking of which, it's the 18 Pro that's finally turned me off the Air. I've been using the Pro for a week now, and realised just how much I've missed those basic features: a better camera, and better battery life. And despite the heft, I'm actually finding it easier to use one-handed, because it's less wide. It's almost a novelty that my thumb can reach the edge of the screen.

It may not be as 'pretty' as the iPhone Air, but the 18 Pro just feels more solid as a device that fits into my life. I'll miss looking at the Air, but it's time to admit: looks aren't everything.

Given the high risk, the temptation to use generative AI in filmmaking is entirely understandable.​ It can bring entire projects to life in a fraction of the time and on a much smaller budget. But at what cost?

The risk of casting AI

AI is capable of throwing together a semi-coherent script and generating visuals that pass the eye test without much scrutiny. Yet, we’re nowhere close to the point where AI can write scenes, organize productions, act naturally, and shoot movies without any human assistance.

The industry will continue to rely heavily on the creativity and vision of its writers, performers, directors, designers, and wider production teams for the foreseeable future, and they’ve already made it abundantly clear that they’re unwilling to share a set with AI if it’s given a starring role. Preventing AI from taking creative jobs in the field was one of the primary causes of the 2023 SAG-AFTRA strike, which caused losses in excess of $5 billion.

Audiences share that sentiment, regularly panning the use of generative AI for creative purposes. And not just heavily AI-generated projects, such as Misaligned, starring AI actress Tilly Norwood, but movies that used it sparingly, such as Young Washington.

While the technology is constantly improving, algorithms are infamous for producing inconsistent narratives, repetitive tropes, and uncanny imagery. With that in mind, something as simple as using generative AI in a promotional poster can, and has, caused backlash. The actual film might be perfectly watchable, but like a poor critic score, that label alone is bound to put plenty of people off buying a ticket.

AI’s role in Hollywood

There’s an important distinction, however. Netflix’s The Gentleman received positive reviews from critics and audiences – despite the use of AI.

The difference is, the technology was used exclusively behind the scenes to break down the series’ action-packed script, inform shooting decisions, and help the production run smoothly. Viewers weren’t presented with AI-generated visuals or a storyline shaped by algorithms. Many productions now use AI in this way, assigning it menial, repetitive pre-production tasks such as script breakdowns, shooting schedules, and budget planning that do not influence what ends up on the screen in any way.

That isn’t a change that’s going to result in another costly strike. I worked as a director’s assistant in Hollywood, and I can confidently say the vast majority of the industry’s workforce would be happy handing over these tasks. Film production is a gruelling, high-stress environment, where 12-hour days are the norm and 16-hour days aren’t uncommon. Nobody suffers through that because they love combing through call sheets for clashes or calculating expenses item by item. These tasks are all vitally important to a successful production, yet creatives would much rather spend their time on the creative side of the process.

The benefits of boring automation

AI, on the other hand, is perfectly happy taking on tedious pre-production processes. And, unlike the creative process, it’s far better at it than humans. It can process far greater amounts of information at far greater speeds, and it doesn’t make mistakes due to fatigue or boredom like humans often do.

Pre-production on a Hollywood film typically takes anywhere from three to seven months. A simple scene breakdown would take days. However, using AI, you can produce the same document in minutes, and the industry has already saved over 5.2M hours of manual work by automating these processes.

Automating ‘boring’ pre-production tasks might not save as much as replacing A-listers with AI stars or generating VFX instead of creating it. However, pre-production still accounts for 10-15% of the total budget on average. That alone would have shaved $80M off of Star Wars: The Force Awakens’ $537M production cost. It’s a huge saving for studios, without the risk of losing your audience.

For AI to pay off, however, the aim shouldn’t be to reduce budgets. Studios should view it as an opportunity to focus more money, time, and attention on the things that actually matter to audiences: refining scripts, improving special effects, perfecting performances. Essentially, creating better films. After all, the real issue with AI in Hollywood isn’t the technology itself – automation has been streamlining creative processes for decades, without complaint. The concern is what happens when all those cost cuts and shrinking teams ultimately compromise the quality of the final cut.

Discover more about Filmustage.

Yet since then, we've had a sequel, Carrie 2 in 1999, plus three remakes: a TV movie in 2002, Kimberly Peirce's 2013 movie, and now an eight-part series released on Amazon Prime Video this week. All of them felt a bit unnecessary, but none prompted as much debate as the latest interpretation.

Warning: this article may contain spoilers for those who have not watched Carrie on Amazon Prime.

Developed by Mike Flanagan (The Life of Chuck), Amazon Prime's Carrie stars Summer H Howell as the eponymous teenage girl with telekinetic powers.

In contrast to the movies, it sticks closer to the documentary format of the novel. It also recognises the impact of the social changes that could make the story worth retelling: metal detectors in US schools, the additional harm of cyberbullying and online harassment via social media.

But it makes other changes that go far beyond merely updating the story to a different age. Carrie's personality has been changed dramatically, making her more optimistic and likeable.

For some, this reduces the concept of what defines Carrie to a series of setpiece scenes. Over on X, someone's provided a side-by-side comparison of one of the most famous: the prom scene, comparing the 1976 Carrie against Amazon's series. Opinions are not favourable to the latter.

"The sound design. The cinematography. The lighting. The editing. The pacing. The score. The performances. How did we regress this badly?" one person asks on seeing the comparison above.

Some point to the inanity of a script that has fleeing characters redundantly explain that the doors are locked and that Carrie's coming for them. Yeah, I think we could gather that.

Others feel that the more sympathetic portrayal of Carrie lightens or even glamourises the scene, making her a heroine in a "crush the bullies" fantasy. The blood looks carefully painted, more like a Halloween costume than a revolting humiliation. Carrie looks like she's "aura farming", one person thinks.

"Carrie is not a 'girlboss moment', it's a mass killing," another person points out. "It ought to convey the horror of a school shooting, hundreds of people trapped in a burning building."

The ending is drastically different too, adding the kind of plot twist that's obligatory in any TV series that's aiming for a second season. I don't want to give too many spoilers away, but it seems to open the possibility of Carrie becoming a kind of X-men/Heroes/The Tomorrow People story.

Flanagan has said that he felt there was more of Carrie's story to tell and that he aims to bring in elements that reflect the larger world that King built through his later books like The Shining, Doctor Sleep and The Institute.

But is the series really telling more of Carrie's story of writing a completely different one that happens to have the same name?

On the outside, it’s the same old Studio we’ve seen for generations, but Apple keeps piling on the improvements and innovations in its silicon, and that’s lead to a machine that’s pulling ahead at the top of many of Creative Bloq’s benchmarking charts, will be a great alternative to the best laptop for animation and is looking hard to catch.

Key specifications

Design, build and display

• Small for a desktop Mac
• Lightweight

Opening up any new Apple product is always an exciting experience as they’re just so well packaged. You pull tabs and tear strips away from the box, then follow arrows to fold down the sides and reveal, in the case of the Studio, the aluminium monolith inside.

The M5 Max version of the Studio looks little different to the way it has all the way back to 2022, with all the changes concentrated on the inside. This means you get something that looks like a Mac Mini under a magnifying glass, only without any of the slight annoyances that have crept into the little computer over the years. The power button is at the back but you can reach it without turning the Mac over, there are loads of USB ports including some Type-As, there's an actual light on the front to tell you if it’s on, and a port for everything so there's less need for a dock or adapter. If you’re lucky enough to own an Apple Studio Display, it will sit on the foot like it was designed to go there.

If you’ve bought the M5 Ultra version of the Mac Studio, it’s heavier than either M5 Max edition due to the presence of an enhanced cooling system for the hotter chip, but the M5 Max is barely heavier than many laptops, and for something that’s settled nicely into the Mac Pro’s niche at the top of the product stack, that’s remarkable.

Design score: 4/5

Features

• Thunderbolt 5 aplenty
• Easy to reach ports

As Apple is reusing the same Studio chassis we’ve seen previously, and the M5 Max version even has the same IO as the M4 Max we reviewed previously, there's very little new here. The Wi-Fi has had a bump to version 7, the Bluetooth to version 6, but there's still plain Thunderbolt 5 (which is plenty fast enough) and a boring old HDMI port.

This machine doesn’t use the M6 - we haven’t seen that chip yet - but the M5 Max takes the already excellent M5 and just adds more to it. Apple is claiming 3.9x faster AI performance than the prior generation and 4.3x the peak AI compute performance of an M3 Ultra. You can spec it with up to 512GB of RAM, if you dare, and you’ll find that this has 50% more memory bandwidth. If that’s not enough, you can use Thunderbolt to connect multiple Mac Studios together to load the most demanding frontier-class open-weight models available. A cluster of four Mac Studios, so it’s claimed, can deliver up to 3x faster AI inference than a single system.

Apple is really pushing the AI capabilities, presumably getting in on the local model and agent action before the first crop of Nvidia RTX Spark machines arrives. But you know what else likes huge amounts of RAM and enormous bandwidth, not to mention loads of processing power? Creative work, that’s what. If you want hardware-accelerated H.264, HEVC, ProRes and AV1 decode, or to colour-grade uncompressed 8K footage while processing multiple concurrent video streams, this is the Mac to do it for you.

It’s also worth noting that Apple has moved its in-house creative apps to a subscription model, called Creator Studio. You get Final Cut Pro, Logic Pro, Pixelmator Pro and a bit more for a monthly fee rather than a large up-front cost, something that will hopefully rekindle interest in these underappreciated apps.

Feature score: 4/5

Benchmark scores

We test every one of our laptops using the same benchmarking software suite to give you a thorough overview of its suitability for creatives of all disciplines and levels. This includes:

• Geekbench: Tests the CPU for single-core and multi-core power, and the GPU for the system's potential for gaming, image processing, or video editing. Geekbench AI tests the CPU and GPU on a variety of AI-powered and AI-boosted tasks.
• Cinebench: Tests the CPU and GPU's ability to run Cinema 4D and Redshift.
• UL Procyon: Uses UL Solutions' Procyon software suite to test the system's ability for AI image generation in Stable Diffusion, its Microsoft Office performance and its battery life in a looping video test.
• Topaz Video AI: We use Topaz Video AI to test the system's ability to upscale video and convert video to slow-motion.
• PugetBench for Creators: We use the PugetBench for Creators benchmarking suite to test the system's ability to run several key tasks in Photoshop and Adobe Premiere Pro, as well as its performance when encoding/transcoding video.
• ON1 Resize AI: Tests the system's ability to resize 5 photos to 200% in a batch process. We take the total time taken to resize the images and divide by 5.

Performance

• Good at everything
• Even the fan isn't too loud

The M5 Max Studio we’ve got here is the top model, with six Super cores, 12 Performance cores (for a total of 18) and a 40-core GPU. It sits one level below the M5 Ultra, which is two M5 Max chips stuck together and can have up to 36 cores, split between 12 Super and 24 Performance, but which costs a lot more. You don’t get Efficient cores on this level of Mac, and it’s essentially the same chip we saw in the M5 Max version of the MacBook Pro, just given slightly more room to breathe in the Studio chassis. It’s backed up in the model sent to Creative Bloq by 128GB of RAM and a 4TB SSD, which contributes to the cost as well as appealing to the new crop of AI types who want to run their models locally.

And so we see it head to the top of our benchmark charts. In Geekbench 6 it takes the top spot in both the single- and multi-core tests, butting even more of a gap between the M5 and its closest rival, the Snapdragon X2 Elite. In Geekbench 2026 it leads the Core Ultra 9 290HX+ in the Razer Blade 18 by almost 3000 points in multi-core rendering, and it’s the first CPU to break the 200-point barrier in Cinebench 2024’s single-core test, in which we used to say that 100 was a good score.

The GPU does well too, sitting 4.2% behind the GeForce RTX 5070Ti in the latest HP Omen 35L - a full-size desktop card - in Cinebench 2026. It beats the popular 24-core Intel Core Ultra 9 275HX laptop CPU in multi-core rendering, and almost doubles the score of the Omen 35L’s CPU, the AMD Ryzen 7 9800X3D beloved of gamers, in the same test.

Speaking of games, if you were thinking of dropping this much money on a Mac and want to use it out of hours, it can get 60FPS out of Cyberpunk 2077 at 4K resolution using the ray-tracing ultra preset, as long as you turn on MetalFX upscaling and AMD FSR frame generation. Drop the res down to 1440p and you can go as high as 89FPS.

In single-precision AI workloads the M5 Max sits between gaming laptops with GeForce chips in our charts, still behind the big desktop PCs. It drops in the Quantized test, however, sitting behind the likes of the Geekom A9 Max, but still ahead of the vanilla M5.

It shines in creative apps though, taking a long lead at the top of the Premiere Pro benchmark chart, beating its portable sibling and the HP Omen 35L. It’s the same in Photoshop, where its results are 6% better than the MacBook Pro with the same CPU.

Performance score: 5/5

Price

In the spec we’ve reviewed here, the Mac Studio costs £6,899/$6,899, but with a lesser spec (M5 Max 16-core, 36GB RAM, 512GB storage) it starts at a more reasonable £2,499/$2,499. That’s still a big chunk of change, even if you are getting a lot of power in a small chassis, and you should definitely look at what a (much larger and noisier) PC with an Nvidia card can do for you in the same price bracket. There's not much else like the Mac Studio, however, and if it has the capabilities you need in a form factor you covet, you’re going to have to open your wallet.

Value score: 4/5

Who is it for?

• Creators and developers

Apple is going big on AI, but has also made a Mac that should be able to do the heavy lifting for many different creative disciplines. It will be overkill for many, as the basic M5 and the M6 that’s on the cusp of arriving are good enough for a lot of tasks. But if you find that you could do with a speed boost in your work, then it’s well worth looking at.

Buy it if...

• Your M1 machine is looking tired
• You prefer MacOS to Windows
• You can’t get enough raw power

Don't buy it if...

• Something cheaper will do the job
• You have little interest in AI
• You’re waiting to see what M6 can do

Also consider

If you're not sure which type of machine is best for you, see our guide to the best 3D printers based on our own reviews, but we rate this mid-range FDM printer very highly, and particularly for beginners who want to avoid buying a printer that they might outgrow too quickly. It provides native multicolour printing and larger print sizes than cheaper options, providing more flexibility to develop your craft.

• Browse 3D printer deals at Amazon

The best multicolour 3D printer for beginners?

The Kobra X isn't the cheapest 3D printer for beginners, but we think it could be a better long-term investment for newcomers who think they're going to want to stick with the craft and possibly even develop professionally. It's simple enough for a newcomer to get started with it but also versatile enough that most users won't outgrow it quickly.

Auto levelling, app control and AI monitoring make it easy to use, while the native 4-colour printing – expandable to 19 – mixed-material capabilities and large build volume mean you can grow into the hobby without having to upgrade your machine a year or so down the line.

If it's your first time looking for a 3D printer, be aware that there are two main families of device: resin and FDM printers.

For most beginners, and for 3D printer users in general, FDM is the way to go. It's easier and safer to handle, has lower running costs, less clean up, and you can get larger print sizes for your money. The exception would be if you really need the extra detail that resin 3D printing can provide for small things like tabletop gaming miniatures or small, detailed jewellery (for that, I'd recommend the Anycubic Photon Mono M7 Pro, currently reduced from $669 to $449 at Amazon).

You can see more deals below.

While Adidas uses several iconic logo designs, its three-stripes have become synonymous with the brand. The athletics company claims that White Fox's four-stripe design is "deceptively similar" to its own, calling for a legal resolution.

White Fox's four-stripe clothing designs have appeared throughout its advertising, which Adidas claims shows an "intention to trade off our reputation". Adidas has trademarked the iconic three stripes in Australia since 1957, and the athletic giant alleges White Fox's designs are "substantially identical with or otherwise deceptively similar."

White Fox's legal team has defended the move, claiming that other Australian brands like Cotton On and Sportscraft sell items with "similar decorative stripes". But Adidas has countered the defence, claiming that White Fox's advertisements feature "genuine" Adidas items, which could form "an association in the minds of consumers."

Adidas is seeking legal damages for the alleged infringement, including profits earned from the products, along with interest and legal costs. For more design disputes, take a look at why New Balance sued this rival athletics brand over a shoe logo or check out the latest Washington Commanders logo controversy.

It's also still a joy, if you'll forgive me. Everyone I know now gets a card I make personally, which is handy when you forget a birthday and can't get to the shop in time (that's happened more than once).

Below, I test and compare the older Joy and the new Joy 2 to see if the updated model is better, especially as right now the original cutter is on sale. I'll start with a basic comparison and then dig a little deeper into how they are to use.

Cricut Joy 2 vs Cricut Joy: quick comparison

The original Cricut Joy is small, simple, and good for cards, labels, and decals, with Smart Materials and foil transfer on board. No Print Then Cut, no scoring, and you have to open the front section every time you use it.

By comparison, the new Joy 2 follows the same small, compact craft-machine idea, but in a lighter, flatter body that's ready to go straight out of the box. Print Then Cut means full-colour stickers, and the bundle includes a scoring tool. This newbie is fully automatic, too, making it incredibly easy to use.

Cricut Joy 2 vs Cricut Joy: price

The Joy 2 starts at $99 / £99.99; for comparison, the original Joy started at $179, so the new model already looks like a bargain. That's for the machine alone, and most people will end up buying a bundle. The Essentials Bundle includes a fine point blade, a scoring tool, a dual-tip pen, a LightGrip mat, a card mat, a portable trimmer, a weeding tool, a scraper and a spatula, so there's plenty to get started with, and you don't need to buy extras before your first project.

These days, the original Joy is much cheaper and often on sale as Cricut promotes its new machine, so expect to pay closer to £50-£69 for this older model. If you are buying a Joy 2, look for bundle deals on the Cricut store for a good deal.

Cricut Joy 2 vs Cricut Joy: design

When the older Joy is in use, you have to open the front section, and it measures 23cm deep. The Joy 2 is 15cm deep and ready to go because nothing folds out. It's slightly wider at 24cm vs 22cm, but much flatter at about 7cm tall vs about 11cm, and it weighs around 1kg vs the original's 1.7kg. It's noticeably easier to pick up.

I used to store my Joy in the cupboard because it was taking up space on my desk. The Joy 2 can just stay out, or you can slide it into a desk drawer. The plug also comes with adaptors for each country, and they're easy to swap, so it travels well (and I've seen people unpack this at airport security, so it's a gadget some crafters can't live without).

Cricut Joy 2 vs Cricut Joy: what's new?

Print Then Cut is the big addition. You print your images on an inkjet printer, upload them to Cricut Design Space, and the Joy 2 cuts them out with high accuracy. The original Joy couldn't do this, so stickers and other full-colour printable projects are new territory.

Scoring is the other one. The Joy 2 comes with a scoring tool, which opens up a lot more projects, especially folded cards. It also comes with a card-making mat, and I've found it ideal for cards. I use it for labels too, cut from card or vinyl.

For quilting, it's been great. I cut cardboard templates on the Joy 2, then use them to cut quilt patterns, and it's a quick, easy way to fit it into a job without bringing out the larger Cricut Maker 5, which I always save for larger projects because of its size.

It's also easier to use than the original Joy. Everything with Cricut is easy, but at first I found it odd to leave the little cutter to decide what to cut and how. It's sensor-driven, so you set up the project in the Cricut app, press your material against the rollers, and the Joy 2 does the rest. If you use the mat, you put it in at the "load and go" stage, and it gets on with it. If you're new to digital crafting, this machine gets you cutting quickly with no hassle.

Cricut says it works with 75+ materials, up from 50+ on the original, and you can use Smart Materials without a mat. Design Space now includes AI tools that make it easier to create highly personalised cards and projects.

Cricut Joy 2 vs Cricut Joy: common features

Both machines are compact and beginner-friendly, work with Smart Materials and foil transfer, and have the same cutting width: 4.5 inches (11.4cm). That's why neither suits big projects. With Smart Materials, you can cut lengths of up to 48 inches (about 122cm) without a mat, which suits long banners and decals.

The software is the same, but when you log in to Design Space, select the Joy 2 from the drop-down menu so you only see compatible projects.

Cricut Joy 2 vs Cricut Joy: which should you buy?

If it's your first compact Cricut, buy the Joy 2. It's well-priced for the tech and use cases, takes up much less room, is easier to use, and adds Print Then Cut and scoring. With all this on offer, I wouldn't buy the original new.

If you already own a Joy, it depends on what you make. For simple cards and vinyl labels, the original is still a really good machine (and if that's all you need, it's on sale so at a reasonable price).

But overall, it's hard not to opt for the new Joy 2. If you want stickers, are short on desk space, or are a crafter like me who wants a small machine to support the big one, the Joy 2 is worth the upgrade.

Recently, billboard ads have stepped up, ditching static design and quite literally thinking outside the box. Unnerving and eye-catching in equal measure, the new Other Mommy billboard proves that breaking convention can have a huge impact.

The new billboard seems pretty unassuming at first. Featuring a picture of lead actress Jessica Chastain alongside the film's bold logo, it could easily be seen as just another static design. But give it time, and slowly it begins to come to life.

As people pass by, a single tear begins to drip from her eye, turning her creepy, waxen-faced smile into a far more foreboding image. The design shines in time-lapse, where you can see the slow yet eerie progression of the billboard.

A crying billboard for a horror movie is marketing genius," one X user wrote, adding "nightmares free of charge." Another chimed in, "That billboard is creepy as hell. Horror fans are eating good!" while another added, "That’s the coolest thing ever omg."

For more creepy design inspiration, check out the Other Mommy poster or take a look at the terrifying new You Can See Everything poster.

Continuing our regular series of artist showcases, he briefly introduces three examples of his work below. If you're inspired, see our guides to the best drawing tablets and the best laptops for drawing and digital art.

Goblin Chirurgeon Reimagined

“A reimagining of Magic The Gathering's Goblin Chirurgeon Card, this piece transforms the original goblin with a hacksaw into a dark surgical session.”

Dragon Tamer

“This was created for a book with my mentor Kenny Vo based on the prompt Land of Legends. Exploring the girl’s story resulted in the tattoo ritual depicted here.”

Ethereal Goddess

“This began as a portrait study from reference before evolving into freeform exploration. I pushed the image even further with fantastical elements.”

You can see more of Marc's work on his ArtStation profile

Today's best drawing tablet deals

Inspired to work on your own digital art? See below for deals on drawing tablets that we've reviewed.

This article originally appeared in ImagineFX. Subscribe to ImagineFX to never miss an issue. Print and digital subscriptions are available.

Creative Bloq is now easier to access than ever before with our on-the-go app, which brings you all the content you know and love from our website, but in a super-streamlined design.
Download the Creative Bloq app for iOS
Download the Creative Bloq app for Android

Little known until a few days ago, this feature buried in customers' account settings reveals some of the information the online retail giant thinks it's gleaned from your activity. This can vary from mundane records about preferences to surprisingly specific and descriptive observations about customers' habits, interests, lifestyles and buttocks.

Amazon introduced About You in May, but the recent Threads post above is largely responsible for bringing it to public attention in recent days. The section lists assumptions that Amazon's AI has formed based on customers' purchasing history, reviews, wishlists and conversations with Alexa (the OP above later concluded that the comment on her physique may have been deduced from her purchase of some “butt scrunch leggings”).

According to Amazon, the assumptions listed can influence the ads and product recommendations that it shows the user. You can edit or remove details that are wrong or that you don't want to be used for recommendations, and you can also import data from other AI bots to help Amazon build up a more detailed profile for you, should you be inclined to do that.

Discovery of the feature has triggered a wave of other Amazon users to check their About You sections and post screenshots of them to social media. In some cases, people have been Amazon's profiling to be eerily accurate, while others have found assumptions that were wildly off the mark.

We know by now that most tech companies profile us. What's surprised people is that Amazon's now letting us read the notes without having to make a freedom of information request. Google and Meta allow users to edit their ad preferences, but they use fairly broad categories. Amazon's go much more granular. It thinks it knows whether you "engage in serious coffee brewing", whether you care for a cat and whether you have "tall books".

While some people are saying they find this algorithmic profiling sinister, most seem to be amused. Perhaps we've become desensitised to data collection amid the normalisation of AI, or maybe Amazon's made a smart move, gaining trust by revealing some of the logic behind its algorithms.

But is Amazon showing us what it knows about us, or just what it wants us to know it knows? The company doesn't claim that the About You page represents all the data associated with an account, but rather a way to view and edit details used to personalise shopping recommendations. It's potentially only a carefully curated view of a small subset of the data it has on us.

The original game was also considered groundbreaking back in 1997 with its varied NPCs and emergent gameplay. It was also a departure in that it was game where you didn't necessarily need to do anything: you could enjoy just driving around and looking at things.

Patrick Kerr, who was a programmer at DMA Design before it became Rockstar North, has now recreated his 2D physics simulation that became the basis for the game's vehicle system, and you can play with it online.

Kerr says the simulation began as a basic wireframe demo written in GFA BASIC on an Atari ST. It was later ported to C for the game. He's now recreated it in JavaScript with sleeker visual presentation.

Car (1) is the default mode, but you can also select Ship (2) or Brick (3) by pressing the respective number keys. Press B to toggle the side collision barriers and G to turn vertical gravity on or off for Ship and Brick modes. Z toggles the camera's dead zone, and X toggles the speed-based zoom just like in GTA.

For developers interested, Kerr's explained how the system works in a post on his website. You can play the demo at Motion Lab.

It's something that's apparent from the moment I press the button that begins my demo, where even the tutorial feels so flashy and cinematic that you could easily imagine being a QTE in one of Sony's first-party blockbusters, except the difference is you're still in control of these actions yourself.

While I'm not able to go into the details of the opening sequences and how they relate to the story of your protagonist Soul, a skilled assassin who has just 66 days to live, while also framed for the murder of his master, I came away feeling like I had experienced a playable kung fu movie, reliving the iconic martial arts duels of classic wuxia films like Once Upon A Time In China or Crouching Tiger, Hidden Dragon, but where instead of just hitting a button prompt, I was in fluid control of my actions.

Mechanically, Phantom Blade Zero may share similarities with other action RPGs or Soulslikes, including colour-coded attacks to indicate whether you should dodge or parry, but what really sells the martial arts fantasy is just how fluid and dynamic the combat animations feel.

It's something you can tell when just holding L1 to block, which in another game would just be quite rigid as your character holds their weapon or shield to guard, but here it looks like Soul is attacking by clashing his blade against his opponent's, precisely how it looks in a wuxia fight.

That authenticity is down to how S-Game has not only mocapped the game's action with real martial artists but even employs wires just like those used in wuxia films that allow actors to perform larger-than-life acrobatics. Even Hong Kong action film veterans including action choreographer and film director Kenji Tanigaki and martial arts legend Donnie Yen are involved, though even in this considerably chunky hands-on time I didn't manage to run into the latter's in-game character.

Yet while these very cinematic animations and visuals realised in Unreal Engine 5 is the main draw, S-Game also elegantly incorporates hand-drawn aesthetics into other parts of the game, perhaps in reference to the original 2D Rainblood: Town of Death, which Phantom Blade Zero is technically a AAA reimagining of.

There's the game map for the Wulin, which is presented in the style of a Chinese ink painting, with more elements added the more of its semi-open world that you explore. While this style means it's actually more of an artistic representation than a more no-frills top-down map, it actually still felt very accurate as I was able to make out spaces and landmarks of areas I had already been to, from its treacherous rocky mountain paths to an outpost held by another faction calling themselves The Mandate.

While this outpost also happens to be located in a bamboo forest that's become a common trope amongst wuxia films, Phantom Blade Zero doesn't just stick to tradition, its kungfupunk aesthetic being a way to differentiate itself with taking subversive influences from more contemporary or even Western ideas, from cyberpunk to modern pop and rock music.

One such example is when I reach a tower where the architecture outside looks quite traditional yet inside also contains a furnace with steel pipes, while it also has another unusual design where it's decorated with red threads, symbolic in Chinese for fate, or marriage, though in this case it's meant to serve a more ominous connotation.

The most kungfupunk example however surely belongs to a fight I got to attempt at a later point in the game, a lion dance grandmaster that's even soundtracked to real percussions, except the lion's head is also designed with metal spikes while it's also able to breathe fire. But while it all looks spectacular, it's definitely not a game that's just dazzling you with cinematic set pieces that require simple QTE prompts.

After all, when I finally reach the Hunter of the Order at a great waterfall for a showdown that the whole chapter had been building up before, it would have been too anticlimactic if I had just wiped the floor with him on the first try.

To some extent, I felt that it might have been tougher than I expected, though it does a very generous thing of at least checkpointing between the phases of a boss fight (difficulty options to make things a little easier, or indeed tougher, are also available). But if Phantom Blade Zero manages to achieve the fine balance of making you look and play like a kung fu master in the same way that Onimusha: Way of the Sword makes you look and play like a samurai master, this could end up being very special.

Phantom Blade Zero is coming to PS5 and PC on 29 October.

Now at version 2026, KeyShot is stronger than ever with a slew of new features to improve the user experience and elevate its capabilities. These include more realistic AI output through AI Shots with the ability to now edit and upscale, an upgraded Move Tool + Custom Pivot, and an all-new gallery feature.

It wouldn't be a software upgrade without some mention of AI, would it? In fact, KeyShot has been using AI for many years, specifically to deliver progressive denoising. But as for AI features, the primary one is AI Shots, which was introduced partway through the 2025 software cycle. KeyShot has determined this is the way to go, devoting a significant amount of development time to improving it.

If you're looking for help on where to start, then you could follow my KeyShot and Photoshop advice for perfect renders every time as well as find out how to create cinematic cameras in KeyShot. To complement KeyShot, you will also need one of the best 3D software packages.

KeyShot 2026: plans and pricing

• KeyShot Professional for $108.25 per month (billed annually)⁣
• KeyShot Business provides seat management features

Most Creative Bloq readers will need a KeyShot Professional subscription. This provides the complete 3D visualisation toolkit for individual designers, freelancers, and small studios. This costs $108.25 per month, although it's billed annually at $1,299.

If you do happen to be a larger team requiring centralised seat management, then you'll have to upgrade to KeyShot Business. For the privilege, this costs $133.25 per month, which equates to $1,599 billed annually.

Value score: 4/5

KeyShot 2026: features

• AI Shots brings generative AI into the heart of the rendering workflow
• A new Gallery window makes it easier to find and organise renderings

The single biggest feature improvement in 2026 is AI Shots. This is KeyShot's generative AI workflow, and it was introduced in KeyShot 2025.2. With it you can either generate images from a CAD model or edit an existing image. Either way, it means artists can utilise generative AI in a dedicated 3D rendering tool rather than relying on alternative tools that were not designed with 3D in mind.

The AI Shots tool is particularly useful for artists wanting to create multiple images with small variations. They can do that easily using the generate multiple images feature when entering the prompt. KeyShot has also improved the overall AI model so it now works with much shorter prompts compared to version 2025.

One of the best improvements of AI Shots is the upscaling feature that enables artists to double or quadruple the resolution of both traditional 3D renderings and AI Shots in a "fraction of the time". This proves more effective than generating an AI shot at a lower resolution and then upscaling.

Beyond AI shots, KeyShot includes a more powerful Custom Pivot + Move Tool. With it users can "give parts or groups multiple custom pivots in addition to origin and centre. These custom pivots can be named for organisational purposes and are stored in your scene."

The move tool gizmo has now also been synchronised between realtime and geometry views, which makes it much more straight forward to work in both views at the same time. It was previously a little confusing, and even though this is only a user experience improvement, it's still a meaningful one.

The final feature to mention is called Gallery, and I love it. It's a new window, and it lets you quickly and easily find your render output once your renderings are complete. It's much more than just a viewer, though. With it you can search, sort, and organise all your rendered output.

Features score: 4.5/5

KeyShot 2026: user interface and experience

• Creating product visualisations in KeyShot is an absolute breeze
• The new AI shots feature provides a generative tool for customising the results

KeyShot is a wonderful piece of visualisation software that takes hardly any time at all to get your head around. Materials, lighting, camera, and environment tools are clearly laid out and easily accessible through various panels and menus. I had no issues finding settings quickly, and with results instantly displayed in the real-time view, it's easy to tweak settings until you get a result you're happy with.

I love how easy it is to cycle through lighting setups and see immediately whether your model catches the reflections nicely or not. With plenty of lighting and material options in the local and cloud libraries artists can quickly translate their 3D model into a beautiful photorealistic visual. Beyond setting up the scene it's also effortless to add post-effects including denoising, sharpening, and bloom, amongst many others.

AI Shots offers a basic interface for adjusting your output using a short prompt. Before doing so, though, you'll have to download near 20GB of AI data, so make sure you've got the necessary storage space before deciding whether KeyShot along with its AI tool is the right visualisation product for you.

I tested AI Shots on my MacBook Pro with 16GB RAM. This is less than the recommended 32GB RAM, but with RAM prices showing no signs of slowing down, I think my laptop is a fair representation of what a lot of users will have access to.

During my tests I found that all elements of the tool, including image generation, upscaling, and element replacement, worked without a hitch. It never crashed, and the progress bar kept moving in the right direction. My only issue with it is that it is quite slow, with some prompts taking upwards of five minutes to process. I found it to be a helpful tool, but it was subject to misinterpreting some of my prompts and couldn't therefore be relied upon for perfect recreation of prompts.

User Experience score: 4.5/5

KeyShot 2026: who is it for?

• KeyShot is designed almost exclusively for product visualisation

KeyShot was built with product visualisation in mind. It offers a simple workflow for 3D artists who regularly find themselves visualising products or product designers who want an in-house solution for creating their own images. KeyShot will work for other types of scenes, but if you're working in film, games, or architecture, then you'll need a dedicated rendering engine such as V-Ray or Cycles (Blender).

Buy it if...

• You create visualisations from CAD
• You need to do this quickly and accurately
• You want a simple rendering solution for small to medium 3D scenes

Don't buy it if...

• You work in film, games, or architecture
• You're regularly rendering large scenes
• You need a rendering engine plugin for your favourite 3D software

And it's the latter that apparently inspired a solo developer to recreate several of Adobe's most popular tools, including Photoshop, Premiere and Illustrator. These were rebuilt using Claude and Rust.

Writing on Hacker News, software developer Brandon Thomas said he was inspired to create the clones after he "got bit by the dark pattern "cancellation fee" one too many times."

"I'm not just some casual vibe coder. I'm a 15-year industry veteran who has written everything from six-nines payment rails that move billions of dollars of transaction volume a day to advanced robotics and optics automation," Thomas announced on Reddit. "I've used Rust extensively for about a decade, and I know how to build and maintain large-scale Rust projects."

"Software is over," he added in a comment on Hacker News.

Is it? Really? It looks to me like Thomas used software to create a clone of some software. And while copying Adobe might sound exciting in a stick-it-to-the-man sort of way, I'm not sure it's a positive omen for the future of software or creativity. Is using AI to create a carbon copy of something as exciting as, say, creating something new?

And it's hard to imagine professional creatives feeling confident enough to place their files, and their livelihoods, in the hands of AI-generated software.

As one Redditor puts it, "There is a pretty fricking huge difference between an unproven software that vaguely looks like Photoshop and probably can do some of the things that Photoshop can do and a dependable, robust productivity software that you can count on working right and being maintained long-term."

More in line with the best horror film posters, the new You Can See Everything poster has all the makings of a spine-chilling scare fest, with suffocating visuals and minimalist design. Bold and frighteningly captivating, I find it impossible to look away, despite how much I want to.

The poster stars the documentary's subject, Elizabeth Holmes, in a claustrophobia-inducing close-up. Spotlighting only half her face, the design forces our gaze to Holmes' almost inhumanly bright blue eye, which appears almost strained in its wideness. Her half-concealed smile only adds to the poster's quasi-uncanny feel, creating palpable discomfort.

In the background, the blurred image of a straight-faced Fielder adds extra tension, while minimalist sans serif font bears the film's title. Slightly cut off on the sides, the title reinforces the subtly suffocating effect, making for a wildly disconcerting design.

"The reflection in the eye is doing more acting than most lead performances this year," one X user wrote in response to the strange poster. "You could tell me this was a horror movie and I'd believe you," another added.

For more poster design, take a look at the new Clayface poster that gets creepier the more you look at it, or check out why horror movie poster design has a problem.

Content warning: this article discusses graphic violence, disturbing imagery, and mature themes in an 18+ game.

I expected a game about pleasure and pain to lean towards the pain. Clive Barker's Hellraiser: Revival leans the other way, and not only because it gets under the skin of the themes and imagery of Clive Barker's The Hellbound Heart (and the first two films). It hides those very adult ideas inside something accessible and familiar, a bit BioShock in its core mechanics, and a bit Condemned in its brutal melee combat and love of leather-bound mannequins, so somewhere around the second hour I stopped noticing the dildos (everywhere) and body parts strewn across the scenery and started enjoying how the systems work.

It's a trick, and a clever one. By the time I realised what I'd been wading through, what I'd just seen as our hero's girlfriend was opened up, torn apart and remade vein by muscle in graphic detail, I was already hooked.

Love in the time of Cenobites

Revival's story is original but inspired by Barker's original book. Set 20 years after the first film in the town of Clairport, north of Detroit. Sunny opens the Genesis Configuration, summons the Cenobites and meets her sticky end. To paraphrase Patrick Stewart, her lover Aidan Lynch sees everything, because they were Netflix and chilling at the time. He makes a deal with the Hell Priest (Pinhead, to most of us), enters the Labyrinth to free her soul, and sets about rebuilding her body.

It's a neat and very games-y setup that challenges Aidan to gather the hearts of three people marked by the box: Charlotte Strauss, an ex-dominatrix known as the Sister of Mercy, and Colin Marlowe, leader of the Scarlet Church, a cult of Leviathan worshippers who want to make Cenobites on Earth. It's a love story of obsession and bad bargains, and feels right at home beside Barker's novella.

So yes, it's very Hellraiser: unapologetically sexual, risqué, bloody and grotesque. Machete swings dissect bodies; past torsos lie on tables, hang like ghoulish wall art, and a late-game ability lets you tear enemies apart with Hellchains that hook and lash in from every direction. It's horrific, often inventively so, the kind of spectacle that demands to be seen and then you kind of want to unsee it, and then look again. So, yep, very Clive Barker (the included digital art book reveals how it was made).

Welcome to the Labyrinth

The game's 'real' world is bold, garish and caricatured, violent and more than a little perverse, but it never lets me forget I'm playing a video game. Then you step through a teasing, glowing crack in a wall into Leviathan's domain, and Unreal Engine 5 drops the cartoon for stark photorealism. That divergence never stopped jolting me and makes the Cenobites' world so much more terrifying: this is the space that looks real, and the real is the space that feels like a game.

The Labyrinth itself is the kind of pain and pleasure Clive Barker would be proud of: puzzling, disorienting and, when the Chatterer (fans know him as the Child) is on my heels, properly terrifying. I found racing down narrow Brutalist concrete canyons, dodging spinning spiked traps and ducking sprouting chains among the best horror I've played this year. Here Aidan is vulnerable and afraid. These escape-meets-walking-sim sections add texture, make you think, and, when you're being hunted, make you feel something genuine.

Outside and back in the sex club, cult mansion and cemetery, Revival dresses up as a light immersive sim, and combat flows freely. A shotgun blast ignites a barrel, fire leaps from one enemy to the next, another steps on a carefully placed bear trap, and the group is dealt with until only body parts remain. The Genesis Configuration piles on top of that. Banishing an enemy with a finisher sends their soul back to the Labyrinth and earns Suffering, a currency you spend on the box's powers. Telekinesis lets you grab sawblades and steel pins from the environment and hurl them at leather-clad cultists, and pyrokinesis lets you pull fire from lamps and shoot it. A revolver, pistol, SMG, assault rifle and a mix of breakable melee knives and bats fill out the earthly arsenal, with upgrades, plus ammo and health you craft from scavenged junk.

It's solid and often a lot of fun, but it stops well short of reinvention. This isn't BioShock, Dishonored or Prey, and it never pretends to be. The Genesis Configuration, for example, never goes as deep as BioShock's Plasmids, though it could really have been developed further and even layered on obsession and pain debuffs and buffs. But no, it's a two-power cube that can occasionally be manipulated as a puzzle to open narrative-dependent gates.

What saves Hellraiser Revival is its structure. The game weaves in and out of Aiden's real world and the Labyrinth, merging the two elegantly in later stages, and the hunt for the three Marked Ones means lots of exploration, Resident Evil-style puzzling and maps that double back on themselves. They're elegantly rendered too, with ornate but grim paintings, statues and reflective surfaces to dwell upon.

It isn't flawless. Combat is at its best when it stays in immersive-sim mode, because boss fights can drag, and controlling Aidan under pressure can feel stiff and unresponsive. When the game asks for subtlety, it fumbles.

Never lets me settle

What lifts Hellraiser Revival is its refusal to sit still. A boss is teased in a stealth section that had me scurrying around a collapsing nightclub while the wraithlike Sister of Mercy hunted me, and then it never happened again. One Labyrinth sequence has you pushing and pulling the environment itself to line up stairs and paths. There's an actual cabin in the woods full of zombies – a fraught survival challenge. And a moment where you have to keep columns erect to hide from Leviathan's Gaze, the beam that exposes people to the darkest parts of their own psyche, is perfectly, gloriously Hellraiser and just on-the-nose enough to be appreciated.

These ideas arrive, get used once, and the game moves on, and that generosity is why the Hellraiser Revival never feels like a retread, even when its bones are familiar.

But those bones are familiar. Scrape away the meaty, fleshy shock value and impressively stylised Labyrinth and underneath is a streamlined immersive sim with very little original in what you do, even if what you experience in its world is refreshingly horrific. And what a world it supports. The Labyrinth is brilliant design, and the Cenobites are horrific and oddly seductive – each scene they're in just feels right. The game embraces its source with total commitment and isn't afraid to get filthy about it, too.

It's partly why the indie devs behind Limerence, a branching narrative thriller about an AI partner decided to explicitly declare on their Steam page that no gen AI has been used in the game's development, including a big bright badge reminiscent of Nintendo's Seal of Quality.

"We found that through Polygon Treehouse, indie dev friends of ours, and they had started this campaign of putting that badge on," Salix Games CEO Jessica Saunders tells me. "We didn't have it originally, and then we were having people messaging me going, 'I just heard about your game, but this has generative AI, so I can't support it.' And I'm like, 'No, it doesn't.' So, we're having to make a big deal out of that."

Yet while AI is being primarily pushed by tech companies, including Epic, which already plans to incorporate Gemini and Claude into Unreal Engine 6, one can see the appeal to create things faster when there are less resources.

Salix Games in particular has felt the pinch of budgets, following its 2019 debut, Dance of Death: Du Lac & Fey, an adventure game with Arthurian characters in Victorian London, which hadn't been financially successful, while the industry post-COVID with layoffs and closures still a daily occurrence, means funding for larger projects has been next-to-non-existent.

But rather than opting for AI shortcuts, Saunders has taken a different approach when considering how to better adapt and survive the current turbulence: "It's like, what could we do for the minimum budget that maximises the skill sets that we have?"

Part of this involved partnering with another studio called Game in a Frame, run by Vincent Joyau, a former colleague of Saunders when she worked at Splash Damage. "He'd been building up his own set of tools for his game, so I asked if I could use them, and he goes, 'I'll do one better, I will use my tools with you and we'll work together to create this new game because we both need to start thinking smaller production-wise if we're going to survive in this industry.'"

Joyau's custom tools created in Unity hasn't just been important for handling Limerence's heavy branching dialogue but also for modelling the game's sole onscreen character, Kai, a 3D AI avatar, who also exists within a fake AI app that's been created with its own fully customisable character creator, outfits, and hairstyles.

This AI partner is only part of the bigger picture of Limerence, which explores how digital dependency can impact our real-world relationships, as the game explores how 30-something woman Claire's life and real-world relationships are impacted by her new-found obsession with a seemingly perfect partner.

This is all presented by the player essentially interacting with Claire's phone, with Saunders citing inspiration from indie games that have played with the premise of snooping on someone's phone, notably A Normal Lost Phone, which then naturally includes other apps that can be accessed, from social media posts, photos, or streamed music.

"It's interesting because these found phone games also use real actors, building up their own photo galleries, but our team lacked production experience, so we've had to pull in a lot of favours from a lot of people," Saunders explains. "We basically said, "Can we have access to your Tik Tok, your Instagram, your small business, and let us use your material, either as a donation or a small fee."

To her surprise, many people were happy to help. They include a friend who loves taking photos of lots of food from restaurants, and a couple of friends who are essentially 'playing' the lead couple in the game, with photos spanning years that show a consistent relationship history that the narrative is able to create a fictional life around.

"We expanded on that too," Saunders says. "We ended up hiring a couple of models through an agency and held a low-budget fake wedding shoot in London which was great fun just to get some of those extra background photos. Honestly, things like Adobe Stock and Unsplash have been great for filling in blanks. That's really allowed us to create a very voyeuristic experience of looking into a real person's phone and just opening up a gallery and seeing things you shouldn't really see."

With stretched resources, mostly going into recording the voice performances, with only Kai's lines being recorded in a sound booth so that the exceptionally clean audio adds to his artificiliaty, it would have been easy for the peripheral aspects of the game to be created with AI, be it arty social media posts or songs playing off a fake streaming app. But Limerence's approach to world-building shows you can make something more lived-in and authentic that also respects and credits artists - you just have to ask.

"The thing is I'm not actually very anti-AI, I think AI as a data tool is marvelous, and it can be very helpful for certain industries," Saunders concludes. "But when it comes to things like making video games, film, TV, and art, I don't see its purpose at the moment because it's just stealing."

Limerence is releasing on PC on 29 October, while Gen's Story, a free demo separate to the final game, is available now on Steam.

Inspired? See our guides to the best game development software and the best laptops for game development.