Author name: Rejus Almole

‘propagation:-paradise-hotel’-review-–-a-pretty-ok-impression-of-‘resident-evil’

‘Propagation: Paradise Hotel’ Review – A Pretty Ok Impression of ‘Resident Evil’

Propagation: Paradise Hotel offers some patently terrifying moments of horror, but between the ever-lingering danger of zombified attacks and a few giant bosses—making for a very Resident Evil-inspired experience—there’s a bit of clunk that tarnishes what could have been a more memorable and cohesive experience. Still, it’s functionally a pretty solid zombie adventure that makes a clear departure from the franchise’s roots as a static wave shooter.

Propagation: Paradise Hotel Details:

Available On:  SteamVR, Quest

Release Date:  May 4th, 2023

Price: $20

Developer: WanadevStudio

Reviewed On:  Quest 2 via PC Link

Gameplay

It’s the zombie apocalypse, and you’re bumming around the bowels of a non-descript hotel in some non-descript part of the world. You won’t have a lot of time to hang with your fatherly security guard pal though because you have to go and find your sister, who is gone for some reason. Okay, so the setup isn’t spectacular, but at least the zombie kill’n is pretty good, right? I’ll lead with an emphatic “yeah, mostly!”

Rule number one of zombies: shoot them in the head. That’s the ironclad directive you’re probably most familiar with, but there are some caveats in Propagation: Paradise Hotel. Shooting zombies in the head multiple times with a pistol makes them very sleepy. No, really. Shoot a zombie three times in the head and they’ll quietly lay down on the ground for a while. Sometime later, usually when you’ve tripped another lurch forward in the narrative, he’ll pop back up at a patently inopportune time to bother you once again.

As clear of a departure from zombie orthodoxy this is, the effect it had on me was something I can’t say I’ve felt in a zombie shooter before. Instead of worrying about walkers popping out from the ceiling (there are a few) or shambolically oozing out from closed doors or windows, you become much more fixated on every single corpse laying in the hallway, of which there are many. You aren’t roaming through an infinite hellscape either, as you’ll be backtracking, learning the layout of the hotel, and tip-toeing around zombies whilst pointing a gun in their face, lest they reawaken and start harassing you again.

Image courtesy WanadevStudio

In effect, any one of them could be waiting for you to lower your guard, open their eyes and grab onto your ankle. Sadly, a preemptive shot in the head is completely ineffectual, which is a letdown in the Immersion department, but more on that below. Just the same, you’ll be cautious because you can’t discount a single corpse, which is a new type of creepy that really kept me on edge. Knowing this, I would have loved the option to cut off some heads to put an end to the contant revisitations, but that’s just not in the cards.

That’s basically the case until you get a shotgun in the latter half of the game, and then those walkers lay down for good because you’ve effectively stumpified their infected brains (finally). What was previously one-on-one battles ramp up to three-on-one battles, putting the game’s only other gun (and most powerful) to the test. You’ll also start to rush through a few new classes of zombie in addition to some more difficult baddies, which offers some interesting variety in difficulty. Will you run into three walkers? One scorpion-style zombie? A ripped dude that can take a ton of shots to the face?

Image courtesy WanadevStudio

While there are a few difficult and unique zombies, there’s really only one true boss in the game. Full disclosure: I disliked it, and while I won’t spoil anything here, rest assured you’ll probably be frustrated too with how to take him down. He is ultra lame, and you’ll want to mute the game just so you don’t have to hear your character constantly shout ad nauseum “I need to knock him out!”

Ok. No. I will spoil the boss. Skip this paragraph if you want to avoid the spoiler: What does “I need to knock him out” even mean?? Don’t I need to kill this bastard? Do I need to knock him out before I kill him? Is that a hint? Do I need something to do that? Maybe I need to call the elevator and rig up something to knock him out? Maybe I need to escape the lobby and head up the stairs to get something I missed? Maybe I need to explode a fire extinguisher in his face to knock him out? Nope. My hand phases right through those, so it can’t be that. Maybe I need to die a dozen times before I learn he has a specific attack pattern with a singular weak point, hit it three times and meander my way to sequel-bait then the end credits? Yup, that’s it.

Anyway, many of the mechanical bits of Propagation: Paradise Hotel are very functional, and work well. The body-based inventory system isn’t overloaded, so you always have what you need, like a medical spray on your left flank, a flashlight that you can clip to your chest or hold in your hand, your 9mm pistol on your right, or the shogun over your shoulder. It’s all there and easy to grab. This compliments a 2D menu that you can pull up, which as mission-essential items, the map, settings, etc.

It’s not easy changing up a user’s expectation of level design when we all know what to expect more or less in a space as familiar as a hotel. Still, the game throws a few curveballs your way to keep you from mechanically looking through every room in the hotel, which spans seven levels. Still, the story itself didn’t feel like a compelling enough driver to keep you moving forward. The found notes add a little flavor, but don’t do enough to flesh out the background of what’s actually going on, leaving you to mostly just bump your head against each task until its complete so you can move onto the next.

Finally, the game, which took me around 3.5 hours to complete, also includes a few puzzles, although all solutions are published in found notes, so you just have to be thorough in your shelf-opening game.

Immersion

Everything about Propagation looks the part, but very little is actually interactable, making it feel more like a flatscreen game than it probably should. There are a few key items you can pick up and use, but everything else is pure set dressing. I don’t want to underplay just how good the game looks, as it offers a visual acuity and variety that makes each room unique, and not at all the sort of copy-paste experience you’d logically expect from a motif that is basically supposed to look extremely uniform. Still, you can’t grab that fire extinguisher, or even pick up a bottle of detergent. You can only open doors and drawers, and interact with keys, key cards, and important notes.

That already feels pretty gamey enough, but just as things start getting good, you grab for an item and a big achievement pops up to ruin the atmosphere—because apparently you need to be constantly reminded that you just collected nine out of 30 secret items. I’d like my full field of view please, since I’m under constant threat of death and everything.

Image captured by Road to VR

One of the big narrative drivers is the game’s found notes, and I generally like the mechanic for its ability to either drive the narrative or unobtrusively flavor its back story. In VR, they can be especially immersive since you’re handling something that’s more of a physical artifact than just a bunch of text on a screen. This is where Propagation fails somewhat, as all notes feature a physical ‘next’ button at the bottom that you have to click, making it more like interacting with an eReader than something that was actually written by someone who lived, survived, and maybe even died in the hotel.

Maybe the notes could be shorter? Maybe they could have used the back of the paper? Maybe a different font? Whatever the case, interacting with a piece of paper shouldn’t feel this unnatural in a VR game.

Image captured by Road to VR

And the wacky unorthodoxy doesn’t stop there. While reloading weapons is a pretty standard experience, the gun’s ammo counter system is definitely not standard. The number indicated isn’t how many bullets you have in the gun, it’s how many are in the magazine. So, if you have 15 bullets in a fresh mag, as soon as you chamber one, the counter says 14.

That’s all well and good for the pistol, but if you forget it when using the shotgun, you may find yourself in deep dog doo-doo as you unintentionally cycle a live round out of the weapon by mistake. Provided you’ve loaded up the shotgun and chambered a round, you may have 3/4 rounds displayed. Once you’re in a tense battle though, and you’re displaying 0/4, you simply can’t be sure whether that 0/4 means you still have one in the chamber, or you don’t. You’ll load back up, shell by shell, until you’re at 4/4 again, but you don’t have any discernible visual indication whether you still have an empty chamber or not, so you cycle the pump just in case. An unspent shell flies out, lands on the floor, and disappears.

While it’s visually interesting and a mostly serviceable shooter despite those inherent flaws, the cherry on the cake is undoubtedly the game’s voice acting, which was clearly farmed out to native French speakers putting on their best American accents. This ranges from “I went to high school in Ohio for a year and picked up the accent pretty well,” to “How do you do, fellow American?” It’s more of an eccentricity than a knock per se, but it leaves me questioning where the hell I am on planet Earth.

Comfort

As a 100 percent walking-based experience that doesn’t include forced locomotion, like on a rollercoaster or similar vehicle, the game proves to be very comfortable, save a single moment when there is some camera shaking. With a wide range of standard comfort options, most everyone will be able to play Propagation: Paradise Hotel without too much issue.

Propagation: Paradise Hotel’ Comfort Settings – May 8th, 2023

Turning
Artificial turning
Snap-turn
Quick-turn
Smooth-turn
Movement
Artificial movement
Teleport-move
Dash-move
Smooth-move
Blinders
Head-based
Controller-based
Swappable movement hand
Posture
Standing mode
Seated mode
Artificial crouch
Real crouch
Accessibility
Subtitles Yes
Languages

English, Italian, German, French, Spanish, Japanese, Korean, Portuguese, Russian, Simplified Chinese, Polish

Dialogue audio Yes
Languages English
Adjustable difficulty
Two hands required
Real crouch required
Hearing required
Adjustable player height

‘Propagation: Paradise Hotel’ Review – A Pretty Ok Impression of ‘Resident Evil’ Read More »

from-michelangelo-to-robots:-this-startup-is-carving-a-new-era-of-sculpture

From Michelangelo to robots: This startup is carving a new era of sculpture

From Michelangelo to robots: This startup is carving a new era of sculpture

Ioanna Lykiardopoulou

Story by

Ioanna Lykiardopoulou

Ioanna is a writer at TNW. She covers the full spectrum of the European tech ecosystem, with a particular interest in startups, sustainabili Ioanna is a writer at TNW. She covers the full spectrum of the European tech ecosystem, with a particular interest in startups, sustainability, green tech, AI, and EU policy. With a background in the humanities, she has a soft spot for social impact-enabling technologies.

Carrara, the Italian city famous for some of the world’s finest marbles, has given birth to the masterpieces of Renaissance masters, such as Michelangelo and Antonio Canova. Now, right in the heart of the city’s quarry district, a startup aims to carve a new era of sculpture in which robots pick up — or rather become — the chisel.

Founded by Filippo Tincolini and Giacomo Massari, Robotor’s mission is to revolutionise the industry by simplifying the sculpting process with the use of robotics and artificial intelligence. The aim is to make sculpture faster, easier, and more sustainable, while enabling artists to create works that would otherwise be inconceivable.

robot creates sculptures of parthenon
Robotor’s machine recreating the Parthenon horse. Credit: Laura Veschi

Drawing from research and the interaction between art, the local area, tradition, and technology, the startup has developed a robot chisel, driven by self-programming software.

It all started in 2004, when Tincolini founded Torart, a company specialising in merging sculpture, contemporary art, and design with the application of new technologies. Torart developed the first robot two years later, and in 2010 Giacomo Massari joined the team as well.

Following increasing orders, the collaboration with world-renowned artists such as Barry X Ball, and the creation of the Arch of Palmyra replica — after the original was destroyed by ISIS in 2015, — the duo founded Robotor in 2019. The size and funding of the company remain undisclosed.

“Our robots are born from sculptors for sculpture,” co-founder Giacomo Massari tells TNW. “They are born from those who perfectly know traditional sculpture in a territory, Carrara, which is the homeland of sculpture.”

Carrara marble
The marble quarries of Carrara. Credit: Laura Veschi

The idea behind their development is “to entrust a robot with tiring work,” Massari adds. “But the entire Robotor project is born from a precise philosophy that is the ‘daughter’ of 18 years of experience in stone processing.”

Putting the robot sculptors to work

The robotic system consists of a mechanical arm that uses electro-spindles of various sizes and capacities, and features an automatic tool change function.

The arm is mounted on a multifunctional, modular base, which contains the necessary electrical and hydraulic components. It’s also accompanied by a seventh-axis rotary table, where the marble and future sculptures stand. The table rotates in full interpolation with the robot’s movements, making the most of its flexibility, and can support a workload of up to 50 tonnes.

robot creates sculpture
Credit: Laura Veschi

At the heart of Robotor’s system is the company’s proprietary software, called OR-OS — designed either for programming by expert operators, or for fully automated self-programming.

The software takes a 3D model and automatically generates an optimised workflow and tool paths without requiring human intervention. It then simulates the process internally to identify any potential issues before sending the instructions to the robotic chisel.

Based on the selected work path, the OR-OS software prompts the robot to choose which types of products and tools it will use for the different stages of the production process: from sculpting, to polishing and cleaning.

robots create sculptures horse of parthenon
Credit: Laura Veschi

The robot then analyses the stone or material it’s going to work on, and reviews its shape and properties, to identify where and how it can be carved, before moving to production.

Besides the robot’s ability to work fully autonomously, users can also opt for programming it and choose the work path it’s going to follow. They’re also able to control various machine functions in real time, including speed adjustment, and monitor the status of the robot and its components, conveniently receiving alerts in case any anomalies occur.

From the human to the robotic hand: speed, accuracy, and innovation

Robotor’s mechanical sculptors come with an array of intriguing advantages.

First off, the robots are designed to eliminate all the tiring phases of the production process, work in extreme environments that would otherwise be endangering humans, and avoid any human error.

“The use of appropriate tools combined with the optimisation of work paths allow for machine use without interruptions 24/7, responding to the need for precise control of work times and costs,” Massari says. This, according to the co-founder, translates into a significant reduction of execution times and expenses.

While cost savings are difficult to quantify, Massari notes that a work of art can, on average, be created by the robots in one-tenth the time that traditional techniques require — excluding manual finishing.

Despite this fast pace, Robotor’s sculptors boast an impressive level of accuracy.

To demonstrate this, Massari mentions the reproduction of The Muse Terpsichore, a famous sculpture completed by Canova in 1811, which was featured in the exhibition “A tempo di danza” at Vetulonia’s Archaeological Civic Museum. The museum opted to display the replica without any hand finishing to avoid confusion between the copy and the original.

Tersicore statue
The Muse of Terpischore, replica by Robotor. Credit: Laura Veschi

Most notably, Massari believes that the robotic chisels can push the boundaries of sculpture, by allowing artists to create works that would have been “inconceivable” before the robots’ facilitation.

“Artists who collaborate with us, such as Quayola and Barry X Ball, present themselves to the public with never-before-seen sculptures designed to be realised with the support of a robotic system, and exploiting its peculiarities,” he says.

“Every work of art is primarily determined by the materials available to the artist and their ability to manipulate them.”

The robots’ diamond-tipped tools that extract material without breaking or splitting it, open new possibilities for material use as well, including conglomerates of stones and earth that can not be processed with traditional methods.

This, according to Massari, is also linked to sustainability, which is accentuated by the software’s 3D model simulation and work path optimisation. That way, the marble block can be excavated with precise knowledge of the final artwork’s dimensions, without any waste.

While suitable for any stone product, Massari says, the robots are mainly targeting three market groups: traditional sculpture workshops aiming to introduce robotics into their production line, artists who wish to manage all creation phases themselves, and designers who produce their own pieces directly. And thanks to their self-programming software, Robotor’s machines are accessible to everyone.

Apart from facilitating the creation of novel artworks, Robotor gains its main satisfaction from the preservation and communication of cultural heritage through the reproduction of great works of the past.

Such an example is the Arch of Palmyra, which was reproduced on a 1:3 scale, based on photographs taken during a documentation campaign of the archaeological site.

Arch of Palmyra replica
The replica of the Arch of Palmyra, displayed in New York. Credit: Laura Veschi

The replica was created in five weeks using 20 tonnes of Egyptian marble, technologically ‘reviving’ the 2,000-year-old monument destroyed by ISIS. In 2016, it was displayed in Trafalgar Square in London, and the City Hall Park in New York.

Another example is Amore e Psyche, one of Canova’s most famous sculptures, completed in 1793 after five years of work, and exhibited at the Louvre Museum.

Eros e psyche Canova Louvre
The original Amore e Psyche by Antonio Canova, displayed at the Louvre Museum. Credit: Louvre Museum.

In 2020, Robotor’s replica was manufactured in just over 10 days, and was featured in the exhibition “Eterna belezza” in Rome, along with 170 other artworks from museums around the world.

Amore e psiche sculpture Robotor
The replica of Amore e Psyche by Robotor. Credit: Laura Veschi

In 2022, the startup made a total of some 400 sculptures. Massari didn’t disclose the price of the machines or the cost for commissioning a sculpture.

Human vs robot: who is the real creator?

Robotor’s mechanical sculptors are fast, accurate, and efficient, and can even rival the works of Renaissance masters. This bids the question: is there enough space for human agency?

Massari’s answer is succinct: the robot will never replace the artist. “It is not a creative, but a mere executor,” he emphasises.

For the co-founder, the robotic chisels are essentially replacing the laborious and time-consuming work that artists would normally entrust to specialised laboratories to execute their vision.

“It is rare to see the artists themselves executing all the phases,” explains Massari, noting that they may intervene in the finishing phase directly, or by giving precise instructions.

“This was how Michelangelo operated, and it is how artists operate today,” he adds.

In fact, he explains that the creative process hasn’t changed at all. Sculptors have used clay or plaster sketches for centuries before moving on to marble work. Today there is one more step: a 3D model that’s created from the sketch, which is essential for the software that controls the robot.

And by replacing the traditional strenuous and time-consuming production process, Robotor aims to actually expand the creative possibilities of artists.

fabio viale sculpture robot
The Dito il Constativo by Fabio Viale. Credit: Laura Veschi

“The history of art is a continuous history of innovations,” says Massari. “Artists have always relied on technological knowledge and ingenuity to find the materials and tools they need to express their dreams, thoughts, visions, or beliefs. And innovation has always had an impact on the way art is made.”

The startup’s ultimate goal is to usher in a new era of sculpture, in which human artists’ expression is supported and expanded through the use of robotics and AI. As the co-founders put it, “this era no longer consists of broken stones, chisels, and dust, but of scanning, point clouds, and design.”

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with


From Michelangelo to robots: This startup is carving a new era of sculpture Read More »

ai-that-decodes-what-mice-see-can-enhance-future-bcis,-say-researchers

AI that decodes what mice see can enhance future BCIs, say researchers

AI that decodes what mice see can enhance future BCIs, say researchers

Thomas Macaulay

Story by

Thomas Macaulay

Senior reporter

Thomas is a senior reporter at TNW. He covers European tech, with a focus on deeptech, startups, and government policy. Thomas is a senior reporter at TNW. He covers European tech, with a focus on deeptech, startups, and government policy.

An AI tool that decodes what mice see could enhance future brain-computer interfaces, according to a new study.

Named CEBRA, the system was developed by researchers at EPFL, a university in Switzerland. Their aim? To uncover hidden relationships between the brain and behaviour.

To test CEBRA (pronounced “zebra”), the team tried to decode what a mouse sees when it watches a video.

“Given the brain is the most complex structure in our universe, it’s the ultimate test for CEBRA.

First, the researchers collected open-access neural data on rodents watching movies. Some of the brain activity had been measured with electrode probes in a mouse’s visual cortex. The remainder came via optical probes of genetically modified mice, which were engineered so their neurons glowed green when activated.

All this data was used to train the base algorithm in CEBRA. As a result, the system learned to map brain activity to specific frames in a video.

Next, the team applied the tool to another mouse that had watched the video. After analysing the data,  CEBRA could accurately predict what the mouse had seen from the brain signals alone.

The team then reconstructed the clip from the neural activity. You can see the results for yourself in the video below:

Unsurprisingly, the researchers aren’t solely interested in the movie-viewing habits of rodents.

“The goal of CEBRA is to uncover structure in complex systems. And, given the brain is the most complex structure in our universe, it’s the ultimate test space for CEBRA,” said EFPL’s Mackenzie Mathis, the study’s principal investigator.

“It can also give us insight into how the brain processes information and could be a platform for discovering new principles in neuroscience by combining data across animals, and even species.”

Nor is CERA limited to neuroscience research. According to Mathis, it can also be applied to numerous datasets involving time or joint information, including animal behaviour and gene-expression data. But perhaps the most exciting application is in brain-computer interfaces (BCIs).

As the movie-loving mice showed, even the primary visual cortex — often considered to underlie only fairly basic visual processing — can be used to decode videos in a BCI style. For the researchers, an obvious next step is using CEBRA to enhance neural decoding in BCIs.

“This work is just one step towards the theoretically-backed algorithms that are needed in neurotechnology to enable high-performance BMIs,” said Mathis.

You can read the full study paper in Nature.

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with


AI that decodes what mice see can enhance future BCIs, say researchers Read More »

a-new-hope:-hv-capital-raises-record-e710m-to-invest-in-european-startups

A new hope: HV Capital raises record €710M to invest in European startups

A new hope: HV Capital raises record €710M to invest in European startups

Linnea Ahlgren

Story by

Linnea Ahlgren

The first quarter of 2023 was pretty bleak for the European startup ecosystem, to say the least. Funding fell a whopping 57% compared to Q1 of 2022, and fundraising is on pace for the lowest total since 2015. As such, the recent developments being heralded from Berlin might bring particularly welcome succour. 

German VC firm HV Capital announced today it has raised its ninth and largest fund ever, with €710 million for investments across all growth phases – all the way from pre-Seed to Series D and beyond. 

The fund is backed mostly by institutional investors from Europe and the US. HV Capital says it will be divided almost evenly into two vehicles: Fund IX Venture and Fund IX Growth. Ticket sizes will range from €500,000 to €60mn.

While the firm will look extensively at deals inside Germany, it also wants to place about 40% of the fund throughout Europe. Reiner Märkle, General Partner at HV Capital, said the record fund would provide the firm with “new opportunities to invest in the next generation of disruptive ideas.”

Indeed, HV Capital, who was an early backer of German e-commerce company Zalando, has already made four investments from the fund. One of these is in Berlin-based SPREAD, who makes augmented engineering intelligence platforms. Another is in GovTech startup Polyteia, also from Berlin, providing authorities with data infrastructure to help “improve and accelerate decision making.” 

Fund IX has also invested in B2B energy management platform ecoplanet, based in Munich, and female-founded monitoring, reporting and verification (MRV) software developer Agreena in Copenhagen, which supports agriculture with regenerative farming practices and carbon monitoring. 

HV Capital said it had established the fund with a view of “advancing ESG in the venture capital ecosystem,” with commitments made under Article 8 of the EU’s Sustainable Finance Disclosure Regulation, or SFDR. 

By the end of the fund’s lifecycle in a decade, the firm says it is targeting at least one-third of women in executive positions across the portfolio. Furthermore, HV Capital will aim to have at least 30% of the fund allocated to companies aligned with the climate goals of the European Investment Fund (EIF). 

SFDR?

If this is the first time you have come across SFDR, consider yourself acquainted with one of the potentially most impactful principles in whether or not your company will receive funding moving forward. Basically, it is a set of rules laid out by the EU designed to counteract greenwashing, and to help investors make more informed decisions about sustainable investment. 

Obligated firms will need to disclose potentially negative consequences an investment decision may have on sustainability factors (environmental and social), and how they are mitigating the impacts, on an annually recurring basis. While it is up to individual member states to decide on financial consequences, there are other potentially adverse effects of non-compliance, such as reputational penalties and sending poor signals to current and future investors.

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with


A new hope: HV Capital raises record €710M to invest in European startups Read More »

uk-competition-watchdog-probes-ai-market-amid-safety-concerns

UK competition watchdog probes AI market amid safety concerns

UK competition watchdog probes AI market amid safety concerns

Siôn Geschwindt

Story by

Siôn Geschwindt

The UK’s competition watchdog has launched a review of the artificial intelligence market, in an effort to weigh up the potential opportunities and risks of a technology Bill Gates dubs as “revolutionary as mobile phones and the Internet.”

The Competition and Markets Authority (CMA) said it would investigate the systems underpinning tools such as ChatGPT in order to evaluate the competition rules and consumer protections that may be required. This, the CMA stated, is to ensure the development and deployment of AI tools is done in a safe, secure, and accountable manner. 

“It’s crucial that the potential benefits of this transformative technology are readily accessible to UK businesses and consumers while people remain protected from issues like false or misleading information,” said CMA chief executive, Sarah Cardell.  

The CMA has set a deadline for views and evidence to be submitted by June 2, with plans to report its findings in September.  

The announcement comes as regulators across the world tighten their grip on the development of generative AI — a technology which can generate text, images, and audio virtually indistinguishable from human output. Hype around this type of AI has been swiftly followed by fears over its impact on jobs, industry, education, privacy — virtually all aspects of daily life.   

In late March, more than 2,000 industry experts and executives in North America — including researchers at DeepMind, computer scientist Yoshua Bengio, and Elon Musk — signed an open letter, calling for a six-month pause in the training of systems more powerful than GPT-4, ChatGPT’s successor. The signatories cautioned that “powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable.”

Meanwhile, Dr Geoffry Hinton, widely referred to as AI’s “godfather,” quit his job at Google this week to talk about the dangers of the technology he helped develop. Hinton fears that generative AI tools could inundate the internet with fake photos, videos, and texts to the extent that an average person won’t be able to “tell what’s true anymore.”

And yesterday, former UK government chief scientific adviser Sir Patrick Vallance told MPs on the Science, Innovation and Technology Committee that AI could have as big an impact on jobs as the industrial revolution.

Anita Schjøll Abildgaard, CEO and Co-founder of Norwegian startup Iris.ai, is optimistic that the probe will address some of these fears and “uphold consumer protections and safely progress the development of AI,” she told TNW. Abildgaard also hopes the review will help address the “competitive imbalance” and “lack of disclosure” present in Big Tech’s proprietary data and training models.

However, while the CMA and many others are clearly concerned about the impacts of AI tools developed by firms such as OpenAI, Microsoft, and Google, Cardell is adamant that the review would not be targeting any specific companies. Rather, she said that the “fact-finding mission” would engage with “a whole host of different interested stakeholders, [including] businesses, academics, and others, to gather a rich and broad set of information”. 

Cardell is also clear that the CMA doesn’t wish to stifle the growth of the rapidly emerging AI industry, but promote it, albeit with a few safeguards. “It’s a technology that has the potential to transform the way businesses compete as well as drive substantial economic growth,” she said. 

A UK government white paper published in March follows a similar trend, signalling ministers’ preference to avoid setting any bespoke rules (or oversight bodies) to govern the uses of AI at this stage. This differs from the EU which is currently in the later stages of finalising its landmark AI Act — the world’s first AI law by a major regulatory body.

While the EU has been first out the gate, according to a new report by the Centre for Data Innovation, politicians should avoid getting swept up in the “hysteria” and shouldn’t “rush to regulate AI before anyone else does because that likely will bode ill, and lead to missed opportunities, for society.”  

Whatever the case may be, the rapid emergence of generative AI has clearly left governments scrambling to figure out if and how to regulate it.

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with


UK competition watchdog probes AI market amid safety concerns Read More »

we’re-closer-than-ever-before-to-creating-a-real-lightsaber

We’re closer than ever before to creating a real lightsaber

Robert Jones

Story by

Robert Jones

Robert Jones is a Post Doctoral Research Associate in the Department of Physics at the Faculty of Natural, Mathematical and Engineering Scie Robert Jones is a Post Doctoral Research Associate in the Department of Physics at the Faculty of Natural, Mathematical and Engineering Sciences of King’s College London. He specialises in theoretical condensed matter physics and its simulation.

Ever dreamed of being a Jedi? Armed with an elegant weapon for a more civilised age, many of us have fantasised about cutting a bright green (or blue) swathe across a galaxy far, far away. Disney even released a patent in 2018 for one such device, although ‘Sword device with retractable, internally illuminated blade,’ doesn’t quite evoke images of cutting through blast doors or droid armour.

That begs the question on this Star Wars Day: is it actually possible to build a lightsaber, and if so, how far away from holding it in our hands and whistling ‘Duel of Fates’ are we? The answer might surprise you —but first, we’ll have to set out exactly what we mean when we say lightsaber and learn some science on the way.

So what would you need to make a lightsaber as we know it?

There are six fundamental boxes a lightsaber must tick: it must light up and glow when in use, it must be able to cut through an object, it must be retractable, it must make a trademark whooshing noise, you must be able to cross them in a fight, and most importantly, it must obey the rule of cool. The bad news is that not all of these are possible at once — yet — but the good news is that all of them are already possible individually — at least, in theory.

Whilst we don’t have access to Kyber Crystals in our galaxy, the laws of physics make more than a fine substitute. The first problem to tackle is the light and cutting edge, and for that, we can utilise the physical principle of laminar flow. This is when all the components in a gas or fluid are moving in the exact same direction without colliding into one another, a little like when you use a shower head. 

This then allows us to use any kind of fluid fuel-oxidiser mix to make a single, high-intensity beam of cutting power. Whilst we may be tempted to use a rocket-grade propellant and fuel, the reality is that something simple like the liquid propane used in BBQs is more than enough. With these ingredients; BBQ fuel, and a laminar flow jet, the task simply becomes one of tuning the fuel mix and valves to get a retractable blade. For the iconic whoosh, it’s a matter of building a circuit with a speaker and accelerometer to make the famed sound of a lightsaber when it’s swung.

Yoda's gorgeous green lightsaber
Yoda’s green plasma blade should only be handled by experts. Credit: JD Hancock

The final touch is the famous colouring of the blade. While Wien’s Displacement law states that the temperature of an object is directly related to its colour, this won’t give us the intense colour we usually associate with this aspect of the Star Wars universe. Instead, by introducing small amounts of specific chemical compounds at the end of the hilt, we can influence the colour. By burning strontium metal for example, we can achieve that iconic Sith red or potassium chloride for the Mace Windu purple. The intense glow of the lightsaber will come from the heat of the plasma generated by the fuel-oxidiser mix.

This still leaves us the issue of being able to cross them in a duel, which necessitates a resistance to temperatures high enough to melt a blast door. At present, the material with the highest melting temperature is the Tantalum hafnium carbide alloy (Ta4HfC5), which melts at a whopping 3990 °C. This unfortunately is the approximate temperature of burning liquid propane. When making something retractable, you also introduce small weaknesses in a metal, which makes fractures and failures more likely. Therefore, even when working with ultra-heat-resistant materials, appropriate care is needed to keep the material from failing due to stress.

This means that any plan to build a lightsaber you can duel with needs to involve not only a heat-resistant material, but a robust one as well. 

How close are we to being able to wield the favoured weapon of the Jedi?

There are two main stumbling blocks to us waving around a screen-accurate lightsaber: the fuel and the duel. Assuming that we’re still following the principle of laminar flow mentioned above, we can achieve a steel-melting, glowing beam by finding a fuel with a high density, and a high burning temperature. The former we want so we can keep the fuel in a nice little rechargeable cylinder, like a battery, and the latter so we can melt through the blast doors of any would-be rebels. 

Acetylene or rocket-grade kerosene could be good candidates, with acetylene being used in plasma cutters and kerosene having put men on the moon in the Apollo programme. Yet, these still don’t quite fit the bill. Acetylene isn’t dense enough to be stored in a battery, and you would need a large tank of it to power a lightsaber for any amount of time. Kerosene on the other hand has a relatively low flame temperature, meaning it would struggle to cut through metal.

Lightsaber colours
Other colours are available. Credit: Remco van der meer

Then comes the issue of being able to cross the blade, as you need a sturdy material which can simultaneously handle the stresses of high temperature and combat with a mortal enemy. My guess for realising this design would be a central core of a high-melting-point material, like Tantalum hafnium carbide alloy, which can be telescopically extended with the high-temperature flame from the propellant and fuel mixture.

The good news is that modern science is making leaps and bounds in this area. Ongoing research into high-density, energy-rich fuels and stress-resilient materials means that we’re now closer than ever to producing a real-life lightsaber. The only question left to ask is, what colour do you want yours in?

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with


We’re closer than ever before to creating a real lightsaber Read More »

german-chip-plant-breaks-ground-in-‘major-step-forward’-for-eu’s-semiconductor-industry

German chip plant breaks ground in ‘major step forward’ for EU’s semiconductor industry

German chip plant breaks ground in ‘major step forward’ for EU’s semiconductor industry

Siôn Geschwindt

Story by

Siôn Geschwindt

Germany’s Infineon has broken ground on its new €5bn semiconductor manufacturing plant in Dresden, less than six months since announcing the project. 

Speaking at the groundbreaking ceremony yesterday, European Commission President Ursula von der Leyen hailed the construction as a “major step forward” toward the bloc’s goal of boosting its market share in semiconductors.

Europe is currently scrambling to reduce its reliance on foreign imports of semiconductor chips from, among others, China, Taiwan, and South Korea. “These are regions where tensions can flare up at any time,” said von der Leyen, in a nod to the current tensions between Beijing and Taipei. “The slightest disruption to trade would immediately hit Europe’s strong industrial base and our internal market hard.” 

The bloc believes that boosting domestic production of the chips — essential components in everything from cars to smartphones — will mitigate the supply shortages that have plagued many of the EU’s vital tech sectors over the past two years. 

Under the EU Chips Act, which got the green light last month, the bloc is mobilising €43bn to double its market share in semiconductors from 10% to 20% by 2030. “The capacity will have to be quadrupled, and this is only possible with companies like Infineon,” von der Leyen said. 

“With this groundbreaking, Infineon is launching an important contribution to the green and digital transformation of our society,” said Infineon CEO Jochen Hanebeck at the ceremony on Tuesday. “Global semiconductor demand will grow strongly and persistently in view of the high demand for renewable energies, data centres and electromobility.”  

A rendering of Infineon’s Smart Power Fab in Dresden which is scheduled for completion in 2026. Credit: Infineon.

Infineon expects production at the plant to start in 2026. The new 300mm fab represents the largest investment in the company’s history, and will add capacity to its current manufacturing site in Dresden.  

Other chip manufacturers are also currently investing in Germany. US-based Wolfspeed is investing €2.7bn to build a plant in Saarland, while Intel is building an even larger factory in Magdeburg that will cost €17bn.

While the EU’s chip plans are taking shape, von der Leyen warned that Europe was still too dependent on raw materials from individual suppliers, pointing to the fact that China controls 76% of the silicon metal supply needed for chip production. 

To secure raw materials closer to home, she said the bloc is exploring new projects in Europe, but also partnerships with countries such as Australia, the US, and Canada. The EU also recently passed the Critical Raw Materials Act which looks to ensure that Europe secures stable supplies of critical rare earth elements and raw materials.

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with


German chip plant breaks ground in ‘major step forward’ for EU’s semiconductor industry Read More »

why-your-next-workplace-training-session-might-be-in-the-belly-of-a-volcano

Why Your Next Workplace Training Session Might Be in the Belly of a Volcano

Whoever said workplace training had to be boring? Thanks to advances in virtual reality, the world of workplace learning and development might finally be something to get excited about…and not a moment too soon.

In the world of training, education, and development, what’s fun and what’s effective are often one and the same. Educators have understood for quite some time that engagement is critical to the learning process. And there’s no quicker way to lose someone’s interest than boring them or lulling them to sleep.

A disengaged student is a student that isn’t learning, and that’s why we find that fun is very much an essential ingredient to the learning process. However, while educators have known this to be true for quite some time, it seems that, in the corporate world, this lesson has yet to truly sink in.

Can you remember the last time you took part in workplace training? If not, that might be because it wasn’t very memorable. If you’re like most people, your experience with workplace training probably amounts to sitting in a drab, windowless room and struggling to keep your eyes open as a grainy video drones at you from a television set in the corner.

Workplace Learning and Development Don’t Have to Be a Drag

Thanks to the emergence of technologies like virtual reality and augmented reality, however, this drab form of corporate training may soon be a thing of the past. Companies are realizing that they can save a considerable amount of time, money, and other resources on training and development. With these portable, cost-effective technologies, organizations can upskill a distributed workforce in a way that’s interactive, inclusive, and sustainable.

In my time working with Gemba, we’ve worked with the companies like Johnson & Johnson, Coca-Cola, and Pfizer to deliver immersive, impactful VR training programs, all while eliminating up to $2M in travel costs and over a ton of CO2  per trainee.

For organizations big and small, that’s a real, meaningful step towards sustainability — and, unlike most other sustainability initiatives, this transition doesn’t entail sacrifice. On the contrary, we’ve found that VR-based training is not only cheaper and more sustainable than traditional forms of training, but it’s actually more effective.

Fun and Efficiency Make VR Training an Inevitability

In our experience helping companies train and develop their employees, we’ve found that the use of VR is associated with significant improvements in educational efficacy.

While many might intuitively understand this, independent research has shown that making the learning process fun or enjoyable not only improves students’ willingness to engage in learning, but also improves their ability to retain information — even if the information itself is dull.

And this phenomenon isn’t unique to students. In fact, we’ve seen the same effect being borne out time and time again in our own work with professionals. An assessment of Gemba’s VR-enabled training with automotive technology supplier, Aptiv, found that a more immersive, engaging, and enjoyable learning experience led to faster, more effective training.

An independent case study of the program, published by the World Economic Forum (WEF),  found that Aptiv was able to reduce what was originally two days of in-person training to just four hours of VR-enabled training. The study concluded that, overall, the use of Gemba’s VR-enabled training allowed Aptiv to upskill its workforce with 80% greater efficiency, compared to real-world training.

And there’s a very real need for more efficient, effective workplace training. According to the Organisation for Economic Cooperation and Development (OECD), over one billion workers globally will need to be upskilled by the year 2030. As technological advancements continue to reconfigure our world, organizations will be hard-pressed to find fast, effective, and scalable means of upskilling such a sizable portion of the workforce.

With VR, The Sky’s No Limit

It’s for these reasons that VR-enabled training’s mainstream adoption is all but an inevitability. Already, leading global enterprises like Unilever, Volvo, L’Oreal, and Nike are making use of VR-enabled training to optimize workforce development. And at the same time, advances in both hardware and software are making virtual and augmented reality platforms more impressive by the minute.

This begs the question — if you can hold a virtual meeting anywhere, why do it in a virtual recreation of some drab, lifeless conference room? Why not in the belly of a volcano? Or a castle floating above the clouds? Or at the bottom of the ocean? Why create a pixel-perfect digital twin of your cramped office space when you can convene anywhere on (or off) Earth?

I know what some of you are thinking — “That sounds rather childish,” or “That seems unprofessional.” To which, I’d respond, is being boring a prerequisite for professionalism?

I’ve had the luxury of working with a wide range of people and professionals throughout my career. I’ve worked in the gaming industry, been a university lecturer, and worked extensively with corporate leadership from around the world. Across all these spaces and demographics I’ve found one thing to be true — people like having fun. As a CTO myself, I can tell you from first-hand experience that you don’t magically become a dullard the moment they put a “C” in front of your title.

And we can expect businesses to be on board as well. Given enjoyment’s proven ability to make learning more effective, you can rest assured that organizations will be eager to inject VR training with all manner of fanciful, gamified fun. And I, for one, can’t wait.

Guest Post


About the Guest Author(s)

Frankie Cavanagh

Frankie Cavanagh

Frankie is a visionary leader, an innovative technologist, and a game-changer in the world of virtual reality and learning. His background in teaching, combined with his passion for game design and XR technology and his skills as an artist, designer, and coder, gives him a unique ability to create immersive and engaging experiences that revolutionize how businesses train and develop their workforce.

Why Your Next Workplace Training Session Might Be in the Belly of a Volcano Read More »

europol’s-operation-spector-leads-to-massive-dark-web-drug-bust

Europol’s Operation SpecTor leads to massive dark web drug bust

Europol’s Operation SpecTor leads to massive dark web drug bust

Linnea Ahlgren

Story by

Linnea Ahlgren

Coordinated by Europol and the US Department of Justice, yesterday, Operation SpecTor seized over €50mn in cash and cryptocurrency, 850 kg of drugs (including amphetamines, cocaine, MDMA and ecstasy pills) and 117 firearms. The latter, contrary to Europol’s choice of emoji, were in all likelihood not water pistols.

🚨 288 dark web vendors arrested in major marketplace seizure.

Operation #SpecTor led to the seizure of:

💶 EUR 50.8 million

💊 850 kg of drugs

🔫 117 firearms

The #MonopolyMarket vendors arrested were also active on other illicit marketplaces.

More ⤵️https://t.co/fDbWJbeFiM pic.twitter.com/086e1XMCUo

— Europol (@Europol) May 2, 2023

The arrested vendors were operating on the marketplace Monopoly Market, selling drugs to customers in exchange for digital currencies. Monopoly has been active since 2019, and one of the candidates hoping for a bigger slice of the darknet pie after law enforcement shut down its highest earner Hydra in April 2022. 

“This operation sends a strong message to criminals on the dark web: international law enforcement has the means and the ability to identify and hold you accountable for your illegal activities, even on the dark web,” Europol’s Executive Director Catherine De Bolle commented

Across Europe, the synchronised arrests took place in the UK (55), Germany (52), the Netherlands (10), Austria (9), France (5), Switzerland (2) and Poland (1). Furthermore, 155 individuals were arrested in the US, and one in Brazil. Europol stated that a number of the suspects arrested were “high-value targets.” 

In addition, authorities obtained buyer lists from the vendors, which could potentially result in the arrests of “thousands of customers” across the globe. 

Transaction data sped up investigations

German law enforcement managed to seize the “criminal infrastructure” supporting the site in 2021. Since then, they have been able to gather intelligence packages that have served as the basis for hundreds of national investigations. 

“The intelligence that Europol shared with us, such as transaction data and virtual currency addresses, helped us to start new investigations and to enrich existing investigations,” the leader of the Dutch team, Nan van de Coevering, told media in the Netherlands. 

In the US, the operation was led by the Department of Justice’s Joint Criminal and Opioid Darknet Enforcement (JCODE) team, set up in 2018. According to US Attorney General Merrick B. Garland, SpecTor “represents the most funds seized and the highest number of arrests in any coordinated international action led by the Justice Department against drug traffickers on the dark web.” 

The 18-month long SpecTor operation is the latest in a row of aptly named high-profile darknet busts by international law enforcement. DisrupTor took place in 2020 with 179 arrests, and Dark HunTor in 2021 with 150 arrests. 

Before it was shut down last year, Russia-based Hydra had about 17 million customers and focused on the trade of illicit drugs, cyberattack tools, forged documents and stolen data. 

Hydra had an estimated revenue of €1.23bn, and German authorities seized €23mn in cryptocurrencies during the bust. Following its shutdown, there was an immediate sector-wide precipitous decline in daily dark web market revenues, from around €4 million to close to €400,000, according to research from Chainalysis.

Closing down Monopoly may not have the same effect, but the ripples will be felt throughout the darknet nonetheless. 

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with


Europol’s Operation SpecTor leads to massive dark web drug bust Read More »

23%-of-jobs-to-be-disrupted-in-the-next-5-years,-wef-predicts

23% of jobs to be disrupted in the next 5 years, WEF predicts

23% of jobs to be disrupted in the next 5 years, WEF predicts

Ioanna Lykiardopoulou

Story by

Ioanna Lykiardopoulou

Ioanna is a writer at TNW. She covers the full spectrum of the European tech ecosystem, with a particular interest in startups, sustainabili Ioanna is a writer at TNW. She covers the full spectrum of the European tech ecosystem, with a particular interest in startups, sustainability, green tech, AI, and EU policy. With a background in the humanities, she has a soft spot for social impact-enabling technologies.

Nearly a quarter of the world’s jobs will be disrupted in the coming five years, the latest report by the World Economic Forum (WEF) has found. Specifically, the job market will experience a 23% churn, as a result of emerging and vanishing positions.

According to WEF, the companies surveyed anticipate that 83m jobs will be lost, albeit offset by the creation of 69m new roles. This still leaves a deficit of 14m eliminated positions, which translates into a 2% contraction of the global workforce.

The report identified three key factors fueling the transformation of the labour market: the green transition, the increased adoption of new technologies, and the slow economic growth alongside the rising cost of living.

Respondents expect that investments that facilitate the green transition of businesses and the broader application of ESG standards will have the strongest net job creation effect, despite a minor displacement percentage. Sustainability Specialists are among the top fastest-growing roles, with Renewable Energy Engineers and Solar Energy Installation and System Engineers flourishing as well.

New technologies are estimated to have an overall positive impact, even though they’ll eliminate positions.

In particular, big data analytics, climate change-environmental management technologies, encryption, and cybersecurity are forecast to be the biggest drivers of job growth. Digital platforms, apps, e-commerce, and AI will also generate more jobs than they eliminate. Only robots will be actually taking our jobs, resulting in a 11.4% role loss.

Correspondingly, AI and Machine Learning Specialists, Business Intelligence Analysts, and Information Security Analysts top the list of the fastest-growing roles. The largest job losses — which are related to the increasing automation and digitisation — are expected in administrative roles and traditional security, factory, and commerce positions. These include, for instance, Cashiers, Bank Tellers, and Accounting, Bookkeeping and Payroll Clerks.

But despite technology’s considerable disruption, respondents believe that the greatest threat to the labour market is the economic downturn. Specifically, the slow economic growth coupled with supply shortages and inflation could be responsible for 87.4% of the net job displacement

From a regional perspective, countries across the world are expected to experience similar levels of disruption in the job market, driven by the same three key factors. The shift is slightly lower in Europe, North America, Middle East, and Northern Africa at 21%, while the highest change is expected in Central Asia at 25%.

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with


23% of jobs to be disrupted in the next 5 years, WEF predicts Read More »

‘break-them-open’-–-new-eu-rules-coming-for-big-tech

‘Break them open’ – new EU rules coming for Big Tech

‘Break them open’ – new EU rules coming for Big Tech

Linnea Ahlgren

Story by

Linnea Ahlgren

As dry and bureaucratic as EU legislation may seem, it can also be groundbreaking and, dare we say it, radical. The bloc has taken a global lead in tackling regulation in areas such as green taxonomy and the much-anticipated AI Act. European lawmakers are also at the forefront in trying to curb the seemingly ever-growing dominance of Big Tech. 

The Digital Markets Act (DMA) is the EU’s tool to attempt to open the digital app marketplace up for smaller competitors. It sets criteria to identify the “gatekeepers” of the market and make them comply with a certain list of do’s and don’ts. 

Among other things, the DMA will promote interoperability, forcing companies like Google, Apple, and Meta to let users link rival apps to their services. This means that Apple will need to release the tightly controlled (and heavily commissioned) grip it exerts through its app store.

In the words of Cédric O, France’s then-digital economy minister, upon the signing of the act last year, “Don’t break them up, break them open.” 

Theoretically, it also means that users of different messaging apps will be able to contact each other from, say, WhatsApp to Telegram, but it is unclear how this would actually be implemented. It will also forbid the gatekeeper companies from doing things such as track their users outside core platforms for targeted marketing without consent. 

While it entered into force on 1 November 2022, the DMA technically began applying yesterday, 2 May 2023. This means that potential gatekeeper tech companies now have until 3 July to notify their core platform services to the European Commission. 

The Commission will then have 45 working days (until 6 September) to decide whether or not they pass the gatekeeper threshold. If the Commission concludes that the company in question does indeed meet the designated criteria, the gatekeeper will then have six months (until 6 March 2024) to comply with the requirements set out in the DMA. 

In the case of non-compliance, the Commission can impose fines of up to 10% of the company’s total worldwide annual turnover. In the event of repeated infringements this can increase to 20% plus periodic penalty payments of up to 5% of the company’s total worldwide daily turnover.

Europe ‘strengthening digital sovereignty’

So who are the “gatekeepers?” According to the DMA, they are platforms in the digital markets that “have a significant impact on the internal market, serve as an important gateway for business users to reach their end users, and which enjoy, or will foreseeably enjoy, an entrenched and durable position.”

As with all legal texts, the criteria go into significant detail. Simplified, they entail that companies will be considered gatekeepers if they have a market capitalisation of more than €75 billion, and 45 million monthly active users in the EU.

There are 10 platform services listed in the DMA. These are: 

  • Online intermediation services;
  • Online search engines;
  • Online social networking services;
  • Video-sharing platform services;
  • Number-independent interpersonal communication services;
  • Operating systems;
  • Cloud computing services;
  • Advertising services;
  • Web browsers;
  • Virtual assistants.

A company may be listed as a gatekeeper for more than one service. 

Together with the Digital Services Act (DSA), the DMA forms one of the central columns of the EU’s digital strategies. They are both part of a regulatory program known as A Europe Fit For the Digital Age.

Adopted three years ago, it is part of the Commission’s ambition to make this Europe‘s ‘Digital Decade’ in which it will “strengthen its digital sovereignty and set standards, rather than following those of others – with a clear focus on data, technology, and infrastructure.”

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with


‘Break them open’ – new EU rules coming for Big Tech Read More »

ai-‘godfather’-quits-google-and-warns-of-dangers-ahead

AI ‘godfather’ quits Google and warns of dangers ahead

AI ‘godfather’ quits Google and warns of dangers ahead

Ioanna Lykiardopoulou

Story by

Ioanna Lykiardopoulou

Ioanna is a writer at TNW. She covers the full spectrum of the European tech ecosystem, with a particular interest in startups, sustainabili Ioanna is a writer at TNW. She covers the full spectrum of the European tech ecosystem, with a particular interest in startups, sustainability, green tech, AI, and EU policy. With a background in the humanities, she has a soft spot for social impact-enabling technologies.

Dr Geoffry Hinton, widely referred to as AI’s “godfather,” has confirmed in an interview with the New York Times that he has quit his job at Google — to talk about the dangers of the technology he helped develop.

Hinton’s pioneering work in neural networks — for which he won the Turing award in 2018 alongside two other university professors — laid the foundations for the current advancement of generative AI.

The lifelong academic and computer scientist joined Google in 2013, after the tech giant spent $44m to acquire a company founded by Hinton and two of his students, Ilya Sutskever (now chief scientist at OpenAI) and Alex Krishevsky. Their neural network system ultimately led to the creation of ChatGPT and Google Bard.

But Hinton has come to partly regret his life’s work, as he told the NYT. “I console myself with the normal excuse: If I hadn’t done it, somebody else would have,” he said. He decided to leave Google so that he could speak freely about the dangers of AI and ensure that his warnings don’t impact the company itself.

In the NYT today, Cade Metz implies that I left Google so that I could criticize Google. Actually, I left so that I could talk about the dangers of AI without considering how this impacts Google. Google has acted very responsibly.

— Geoffrey Hinton (@geoffreyhinton) May 1, 2023

According to the interview, Hinton was prompted by Microsoft’s integration of ChatGPT into its Bing search engine, which he fears will drive tech giants into a potentially unstoppable competition. This could result in an overflow of fake photos, videos, and texts to the extent that an average person won’t be able to “tell what’s true anymore.”

But apart from misinformation, Hinton also voiced concerns about AI’s potential to eliminate jobs and even write and run its own code, as it’s seemingly capable of becoming smarter than humans much earlier than expected.

The more companies improve artificial intelligence without control, the more dangerous it becomes, Hinton believes. “Look at how it was five years ago and how it is now. Take the difference and propagate it forwards. That’s scary.”

The need to control AI development

Geoffry Hinton isn’t alone in expressing fears over AI’s rapid and uncontrolled development.

In late March, more than 2,000 industry experts and executives in North America signed an open letter, calling for a six-month pause in the training of systems more powerful than GPT-4, ChatGPT’s successor.

The signees — including researchers at DeepMind, computer scientist Yoshua Bengio, and Elon Musk — emphasised the need for regulatory policies, cautioning that “powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable.”

Across the Atlantic, ChatGPT’s growth has stirred the efforts of EU and national authorities to efficiently regulate AI’s development without stifling innovation.

Individual member states are trying to oversee the operation of advanced models. For instance, Spain, France, and Italy have opened investigations into ChatGPT over data privacy concerns — with the latter being the first Western country to regulate its use after imposing a temporary ban of the service.

The union as a whole is also moving closer to the adoption of the anticipated AI Act — the world’s first AI law by a major regulatory body. Last week, Members of the European Parliament agreed to advance the draft to the next stage, called trilogue, in which lawmakers and member states will work out the bill’s final details.

According to Margrethe Vestager, the EU’s tech regulation chief, the bloc is likely to agree on the law this year, and businesses could already start considering its implications.

“With these landmark rules, the EU is spearheading the development of new global norms to make sure AI can be trusted. By setting the standards, we can pave the way to ethical technology worldwide and ensure that the EU remains competitive along the way,” Vestager said when the bill was first announced.

Unless regulatory efforts in Europe and the globe are sped up, we might risk repeating the approach of Oppenheimer of which Hinton is now sounding the alarm: “When you see something that is technically sweet, you go ahead and do it and you argue about what to do about it only after you have had your technical success.”

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with


AI ‘godfather’ quits Google and warns of dangers ahead Read More »