No need for more scare stories about the looming automation of the future. Artists, designers, photographers, authors, actors and musicians see little humour left in jokes about AI programs that will one day do their job for less money. That dark dawn is here, they say.
Vast amounts of imaginative output, work made by people in the kind of jobs once assumed to be protected from the threat of technology, have already been captured from the web, to be adapted, merged and anonymised by algorithms for commercial use. But just as GPT-4, the enhanced version of the AI generative text engine, was proudly unveiled last week, artists, writers and regulators have started to fight back in earnest.
“Picture libraries are being scraped for content and huge datasets being amassed right now,” says Isabelle Doran, head of the Association of Photographers. “So if we want to ensure the appreciation of human creativity, we need new ways of tracing content and the protection of smarter laws.”
Collective campaigns, lawsuits, international rules and IT hacks are all being deployed at speed on behalf of the creative industries in an effort, if not to win the battle, at least to “rage, rage against the dying of the light”, in the words of Welsh poet Dylan Thomas.
Poetry may still be a hard nut for AI to crack convincingly, but among the first to face a genuine threat to their livelihoods are photographers and designers. Generative software can produce images at the touch of the button, while sites like the popular NightCafe make “original”, data-derived artwork in response to a few simple verbal prompts. The first line of defence is a growing movement of visual artists and image agencies who are now “opting out” of allowing their work to be farmed by AI software, a process called “data training”. Thousands have posted “Do Not AI” signs on their social media accounts and web galleries as a result.
A software-generated approximation of Nick Cave’s lyrics notably drew the performer’s wrath earlier this year. He called it “a grotesque mockery of what it is to be human”. Not a great review. Meanwhile, AI innovations such as Jukebox are also threatening musicians and composers.
And digital voice-cloning technology is putting real narrators and actors out of regular work. In February, a Texas veteran audiobook narrator called Gary Furlong noticed Apple had been given the right to “use audiobook files for machine learning training and models” in one of his contracts. But the union SAG-AFTRA took up his case. The agency involved, Findaway Voices, now owned by Spotify, has since agreed to call a temporary halt and points to a “revoke” clause in its contracts. But this year Apple brought out its first books narrated by algorithms, a service Google has been offering for two years.
The creeping inevitability of this fresh challenge to artists seems unfair, even to spectators. As the award-winning British author Susie Alegre, a recent victim of AI plagiarism, asks: “Do we really need to find other ways to do things that people enjoy doing anyway? Things that give us a sense of achievement, like writing a poem? Why not replace the things that we don’t enjoy doing?”
Not a fan of AI: singer-songwriter Nick Cave. Photograph: Simona Chioccia/Shutterstock
Alegre, a human rights lawyer and writer based in London, argues that the value of authentic thinking has already been undermined: “If the world is going to put its faith in AI, what’s the point? Pay rates for original work have been massively diminished. This is automated intellectual asset-stripping.”
The truth is that AI incursions into the creative world are just the headline-grabbers. It is fun, after all, to read about a song or an award-winning piece of art dreamed up by computer. Accounts of software innovation in the field of insurance underwriting are less compelling. All the same, scientific efforts to simulate the imagination have always been at the forefront of the push for better AI, precisely because it is so difficult to do. Could software really produce paintings that entrance or stories that engage? So far the answer to both, happily, is “no”. Tone and appropriate emotional register remain hard to fake.
Yet the prospect of valid creative careers is at stake. ChatGPT is just one of the latest AI products, alongside Google’s Bard and Microsoft’s Bing, to have shaken up copyright legislation. Artists and writers who are losing out to AI tend to talk sorrowfully of programmes that “spew rubbish” and “spout out nonsense”, and of a sense of “violation”. This moment of creative jeopardy has arrived with the huge amount of data now available on the web for covert harvesting rather than due to any malevolent push. But its victims are alarmed.
Analysis of the burgeoning problem in February found that the work of designers and illustrators is most vulnerable. Software programs such as Midjourney, Stable Diffusion and DALL.E 2 are creating images in seconds, all culled from a databank of styles and colour palettes. One platform, ArtStation, was reportedly so overwhelmed by anti-AI memes that it requested the labelling of AI artwork.
At the Association of Photographers, Doran has mounted a survey to gauge the scale of the attack. “We have clear evidence that image datasets, which form the basis of these commercial AI generative image content programs, consist of millions of images from public-facing websites taken without permission or payment,” she says. Using the site Have I Been Trained which has access to the Stable Diffusion dataset, her “shocked” members have identified their own images and are mourning the reduction of the worth of their intellectual property.
The opt-out movement is spreading, with tens of millions of artworks and images excluded in the last few weeks. But following the trail is tricky as images are used by clients in altered forms and opt-out clauses can be hard to find. Many photographers are also reporting that their “style” is being mimicked to produce cheaper work. “As these programs are devised to ‘machine learn’, at what point can they generate with ease the style of an established professional photographer and displace the need for their human creativity?” says Doran.
For Alegre, who last month discovered paragraphs of her prize-winning book Freedom to Think were being offered up, uncredited by ChatGPT, there are hidden dangers to simply opting out: “It means you are completely written out of the story, and for a woman that is problematic.”
Alegre’s work is already being misattributed to male authors by AI, so removing it from the equation would compound the error. Databanks can only reflect what they have access to.
“ChatGPT said I did not exist, although it quoted my work. Apart from the damage to my ego, I do exist on the internet, so it felt like a violation,” she says.
“Later it came up with a pretty accurate synopsis of my book, but said the author was some random bloke. And, funnily enough, my book is about the way misinformation twists our worldview. AI content really is about as reliable as checking your horoscope.” She would like to see AI development funding diverted to the search for new legal protections.
Fans of AI may well promise it can help us to better understand the future beyond our intellectual limitations. But for plagiarised artists and writers, it now seems the best hope is that it will teach humans yet again that we should doubt and check everything we see and read.
The Digital4Business consortium, which has several Irish members, aims to launch its first MSc programmes in January 2024.
The National College of Ireland (NCI) helped to launch a new pan-European digital and entrepreneurial skills project that aims to provide a steady pipeline of talent to SMEs in the region.
NCI is one of 15 partners from seven European countries that are taking part in the consortium leading the project, which is called Digital4Business.
Digital4Business is a four-year initiative that will see various EU institutions and businesses work together to devise and deliver a market-led postgraduate programme to help SMEs access a pipeline of digital talent.
Other programmes launching in the coming few months and years will concentrate on key topics such as cloud, data analytics, AI, cybersecurity, blockchain, IoT and quantum computing.
Overall, the project will cost €19.92m. The programmes that will result from it will offer both industry and academic accreditation. The consortium will be focused on the practical application of advanced digital skills within European companies.
The initiative is being funded by the European Commission’s Digital Europe programme, which focuses on the digital transformation of Europe’s society and economy. The funding award to Digital4Business is one of the largest awards the programme has made to date.
Digital4Business was officially launched at an event in the IFSC in Dublin today (21 March).
The project began in December 2022. The consortium aims to launch the first part-time and full-time MSc programmes in January 2024.
Speaking by video link at the event, Minister for Further and Higher Education, Research, Innovation and Science, Simon Harris, TD, highlighted the project’s relevance as part of the European Year of Skills.
“2023 is European Year of Skills – the focus is on helping people get the right education to be prepared for quality jobs, and to address specific skills shortages that businesses are experiencing – particularly SMEs. Digital4Business directly serves this mission.”
Dara Calleary TD, Minister of State for Trade Promotion and Digital Transformation attended the event in person.
“Digital4Business’ focus on the practical application of advanced digital skills within companies, and especially, within our small and medium businesses, is of great importance. This type of talent development is essential to ensure that the skills and the expertise are in place for businesses to maximise their digital potential – to take advantage of the opportunities digital presents and to assist them in maintaining their competitive edge,” he said.
As well as NCI, the other Irish partners involved in Digital4Business are IT company Terawe, Skillnet Ireland and Digital Technology Skills Limited. Digital agency Matrix Internet co-headquartered between Ireland and Belgium is also involved.
10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.
With a click of the right-mouse button, my musclebound barbarian sinks his axe into the ground behind him, sweeps it forward and creates a shock wave that obliterates everything in its path. Ahead, a horde of undead creatures is repulsed by the blast, zombies flayed by the force of the air, skeletons scattered across the ground, wraiths dissipating into spectral dust. The room’s furnishing fly with them, chairs, candlesticks and barrels smashing into the far wall. The ground itself is scarred by the attack, a conical depression left in the floor as if struck by a meteorite airburst.
I’ve performed this attack countless times over the last weekend, and it never fails to light up my brain like Blackpool in September. The Diablo series represents video gaming in its purest and perhaps most reductive form and has exploited these feedback loops to enormous success in the last 25 years, reworking the complex rulesets of role-playing games into something less cerebral and more sensory. While there’s an argument to be had about how intellectually nourishing these games may be, Diablo 4 has a lot of seductive power. Clicking monsters to death in this game feels dangerously good.
Diablo 4 returns to being the moody goth kid of its RPG social group. Photograph: Blizzard
Yet having spent 48 hours with the game during its beta phase, it’s clear there’s more to this than mindless monster-bashing. Diablo 4 sees the series return from a long hiatus after a third game that proved controversial in more ways than one. Partly because of this, it looks both backward and forward, addressing some criticisms of Diablo 3 while striving to compete in a world that has changed dramatically since 2012.
After a mixed reception to the colourful visuals of Diablo 3, Diablo 4 returns to being the moody goth kid of its RPG social group: pale-faced, clad in black and obsessed with death. The opening area, named Fractured Peaks, is an oppressive place where muddy, monster-ravaged villages cling to the edges of a snowy mountain range, with warrens of caves and dungeons concealed beneath the frozen surface. Said dungeons revel in their own dinginess. Painted in abundant dark shades, much like FromSoftware’s Bloodborne, the blackened walls and floors are slick with decaying viscera and often writhe with strange tendrils that grasp at you from the stonework.
Diablo 4’s appeal to the past isn’t purely stylistic. As your character accrues power across the game’s dark fantasy adventure, you must choose how to channel that power, picking skills and abilities that complement one another to make your chosen warrior an unstoppable destructive force. Diablo 4 ditches the previous game’s overly streamlined approach, returning to a more traditional skill tree that shows your character’s entire power trip at a glance.
Players now carve their way through a huge open world. Photograph: Blizzard
I tested two of the five available character classes in the open beta – the barbarian and the sorcerer. What became obvious during my time with them is how intuitive character progression is. My sorceress, for example, offered an array of elemental powers to choose from. I could have made her an incandescent pyromancer, or a weaponised Elsa who froze her enemies to death. Instead, I focused on electrical abilities, Emperor Palpatine-ing my way through dungeons by zapping demons with bouncing bolts of lightning. This wasn’t the limit of my options, either. Diablo 4 let me further tailor these attacks to produce a collectible item known as “Crackling Energy”. As I plucked these orbs of static electricity from fallen foes, they’d discharge automatically when approaching new enemies. Hence, my sorceress could fry whole groups of demons before casting her first spell – a delightful sensation.
Structurally, Diablo 4 is different, as players now carve their way through a huge open world. For the beta, only the Fractured Peaks area was available to explore, but this nonetheless represents a sizeable and impressively freeform area. Although there is a central story to follow, it’s easy to get side-tracked into some offshoot adventure, helping a villager find her missing husband in some shadowy forest or delving into optional dungeons with foreboding names such as the Black Asylum. These secondary activities are tied together by “Renown”, a currency that, when accrued, periodically rewards players with extra gold, skill points, and other bonuses.
The looser structure creates a more coherent world, but it doesn’t radically change how Diablo plays. Instead, the open world exists mainly to facilitate Diablo 4’s new status as a persistent online game. Diablo 4 has extensive multiplayer features, with other players wandering freely around the game world able to periodically fight together as they explore individually, or actively join clans and embark on quests together. This ever-present multiplayer element could prove controversial, but interaction with other players isn’t mandatory, and you can happily plunder dungeons and pursue the central storyline solo.
Cults. Prophecies. More helpless villagers. Photograph: Blizzard
While the story has always been a part of Diablo, its role is small compared with other RPGs – largely an excuse for players to mash monsters by the million. But Diablo 4 makes a more concerted effort to grab the player’s attention, breaking up the action with more elaborate cutscenes and dialogue that dwell on individual characters, and takes more time to explore the game’s pseudo-Christian lore. These sequences bring with them all the flair you’d expect from Blizzard, and an impressive cast that includes veteran voice actors such as Troy Baker and Jennifer Hale, alongside Hollywood names like Ralph Ineson.
Broadly, it’s a typical fantasy adventure, a grand battle between good and evil. There are cults. There are prophecies. There are more helpless villagers than you can shake a pitchfork at. But there is also an attempt at more nuanced characterisation. The main antagonist – the demonic goddess Lilith – is not wholly villainous, while the fallen angel Inarius, a central figure in the religion of the game’s longsuffering humans, is not wholly good. There’s enough of interest to be audible above the sound of battle, and it helps that the game takes itself seriously, avoiding the temptation to lace the narrative with knowing side-glances and ironic gags.
Some questions remain. While Diablo’s character progression is slick and intuitive, will it offer the same level of flexibility as other ARPGs, most notably Path of Exile, which stepped in during Diablo’s long absence? Moreover, what does this new multiplayer structure mean for Blizzard’s long-term monetisation plans – will we eventually be asked to pay for a subscription? It appears inevitable it will continue to evolve after launch, and the question is what form will that evolution take. This is a game that could change shape substantially in the coming years. In its current form at least, Diablo 4 seems like a worthy ascendant to the throne of destruction.
GTC Nvidia’s latest gambit? Entrenching itself as a key part of the semiconductor manufacturing supply chain.
At GTC this week, the chipmaker unveiled cuLitho, a software library designed to accelerate computational lithography workloads used by the likes of TSMC, ASML, and Synopsys, using its GPUs.
The idea behind the platform is to offload and parallelize the complex and computationally expensive process of generating photomasks used by lithography machines to etch nanoscale features, like transistors or wires, into silicon wafers.
“Each chip design is made up of about 100 layers and in total contains trillions of polygons or patterns. Each of these 100 layers are encoded separately into a photomask — a stencil for the design if you will — and, using a rather expensive camera, are successively printed onto silicon,” Vivek Singh, VP of Nvidia’s advanced technology group, explained during a press conference on Monday.
Originally, photomasks were just a negative of the shape engineers were trying to etch into the silicon, but as transistors have gotten smaller these photomasks became more complex to counteract the effects of optical distortion. If unchecked, this distortion can blur these features beyond recognition. This process is called optical proximity correction (OPC) and more recently has evolved into inverse lithography technology (ILT). In the case of the latter, the photomasks look nothing like the feature they’re designed to print.
And the more ornate these photomasks get, the more computational horsepower is required to produce them. However, using GPUs, Nvidia believes it can not only speed up this process, but reduce the power consumption required. The company claims that cuLitho running on its GPUs is roughly 40x faster than existing computational lithography platforms running on general purpose CPUs.
“It’ll help the semiconductor industry continue the pace of innovation that we’ve all come to rely on, and it’ll improve the time to market for all kinds of chips in the future,” Singh claimed.
However, at least in the near term, Nvidia’s expectations seem to be a little more grounded. The company expects fabs using cuLitho could produce 3-5x more photomasks a day while using 9 percent less power, which if true, should help to boost foundries’ already thin margins
And with the likes of ASML, Synopsys, and TSMC lining up to integrate Nvidia’s GPUs and libraries into their software platforms and fabs, we won’t have to wait long to see these claims put to the test.
TSMC is already investigating Nvidia’s GPUs and cuLitho to accelerate ILT photomasks, while ASML and Synopsys are working to integrate support for GPU acceleration using cuLitho in their computational lithography software platforms.
And while Nvidia execs would love to sell its latest and most expensive GPU architectures to these companies, Singh notes that the library is compatible with GPUs going back to the Volta generation, which made its debut in 2017.
While Nvidia is using GPUs to accelerate these workloads, it’s worth noting that cuLitho isn’t using machine learning or AI to optimize semiconductor design just yet. But it’s no secret that Nvidia is also working on that particular problem.
“Much of this has to do with accelerating the underlying primitive operations of computational lithography,” Singh said. “But I will say that AI is very much in the works in cuLitho.”
As our sister site The Next Platformreported last summer, Nvidia has been working on ways to accelerate computational lithography workloads for some time now. In a research paper published in July, engineers at the company used AI to design equivalent circuits 25 percent smaller than those created using traditional EDA platforms.
Nvidia is hardly the only company investigating the use of machine learning to accelerate circuit design. Synopsys and Cadence have both implemented AI technologies into their portfolios, while Google researchers developed a deep-learning model called PRIME to create smaller and faster accelerator designs. And previously, the company used reinforcement learning models to design portions of its tensor processing unit (TPU).
With that said, the addressable market for something like cuLitho isn’t that big, and thanks to efforts by the US Commerce Department to stifle China’s fledgling semiconductor industry, the number is only getting smaller.
cuLitho will almost certainly be subject to US export controls governing the sale of advanced semiconductor manufacturing equipment and software to countries of concern, which for the moment means China. Pressed on this point, Singh said the library would be “available wherever this end-to-end OPC software is available,” but declined to comment further on US trade restrictions. ®
