Amazon Web Services has a new region for its cloudy customers to consider – in Indonesia.
The AWS Asia Pacific (Jakarta) Region has three availability zones, meaning the cloud colossus has created three geographically isolated locations.
Which is a very good idea because Jakarta, Indonesia’s capital city, is sinking. Literally. Land subsidence is such an issue in the city that Indonesia has planned to move its capital to a new location on the island of Borneo.
The Register fancies AWS has either built close to nearby hills, or otherwise satisfied itself that the resilience it desires can be achieved for the expected life of this set of bit barns.
Whatever AWS did to get up and running, it has competition.
Google Cloud is already in Jakarta, and Microsoft has announced a region in Indonesia for its Azure cloud.
Alibaba has had two datacentres in Indonesia since July 2020 and is already planning a third. Tencent has one too, and is planning a second facility in Jakarta.
China’s big clouds see South-East Asia as their most comfortable offshore expansion targets.
AWS’s announcement of its new region advanced its usual optimism about local customers appreciating a low-latency service and in-country storage, and a desire to serve entrepreneurs at all stages of business development, as well as governments. Which is almost exactly what every cloud provider says at the opening of a new region.
The draft DSA measures include tackling illegal content and holding social media platforms accountable for their algorithms.
The European Parliament has voted in favour of the long-debated Digital Services Act (DSA), Europe’s attempt to shift the balance of power from the hands of Big Tech and into the hands of EU residents.
MEPs voted 530 votes to 78 – with 80 abstentions – to approve the draft DSA text that will see tech giants held accountable for content on their platforms in a spate of new rules and regulations.
The move comes just a month after MEPs voted in favour of the Digital Markets Act (DMA), a similar set of proposed laws that seek to impose stricter rules around tech competition in the EU and rein in the monopoly large multinationals hold in Europe’s digital space.
The set of draft DSA measures include tackling illegal content, preventing the spread of misinformation and disinformation and holding social media platforms accountable for their algorithms.
The Parliament introduced several changes to the initial proposal by the European Commission, including exempting small businesses from certain DSA obligations, making targeted advertising more transparent and easier to refuse, and prohibiting targeted ads for minors.
Online platforms will be prohibited from using “deceiving or nudging techniques to influence users’ behaviour through ‘dark patterns’”, according to the revised DSA draft. Large platforms will also be required to provide “at least one recommender system that is not based on profiling”.
The approved text will now be used as the mandate to negotiate with the French presidency of the Council, representing member states. The negotiation will be led by Christel Schaldemose, an MEP from Denmark.
“Today’s vote shows MEPs and EU citizens want an ambitious digital regulation fit for the future,” she said after the vote. “Online platforms have become increasingly important in our daily life, bringing new opportunities, but also new risks,” she added.
Schaldemose said it was the duty of the European Union to make sure “what is illegal offline is illegal online” and that new digital rules need to benefit consumer and citizens, not Big Tech.
She said that, if enacted and enforced correctly, the DSA has the potential to inspire other nations such as the US to take on Big Tech and “safeguard democracy” before it’s too late.
“Every modern disinformation campaign will exploit news media channels on digital platforms by gaming the system,” Haugen told MEPs in her opening statement. “If the DSA makes it illegal for platforms to address these issues, we risk undermining the effectiveness of the law,” she said.
Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.
When FC Barcelona took to the pitch for the 2021 Spanish Super Cup final, the trophy wasn’t the only prize at stake.
Thousands of blaugrana fans were also keeping an eye on the market for FCB’s “fan token”, the club’s very own cryptocurrency. Socios, the web-based platform that pioneered fan tokens, had promised to “burn” 20,000 tokens for every goal Barcelona scored – and 40,000 if they lifted the cup.
In theory, success on the pitch would increase the scarcity of the currency, boosting its value. In practice, Barcelona lost the game and, footballing passions aside, it didn’t make much difference anyway. With 3.5m of the tokens in circulation, not to mention millions more retained by the club for future issuance, a few thousand here or there wouldn’t have moved the needle.
Football finance expert Kieran Maguire thinks clubs have latched on to crypto because revenues from other sources are starting to level off, having risen reliably for decades.
“Football clubs have realised that we’re now at max broadcast revenues, with modest growth at most to look forward to,” he said.
“As far as commercial sponsors are concerned, we’re seeing deals being renewed but not with increased money. The only way to increase matchday sales is to increase prices and fans are reluctant.”
Manchester United – whether one believes the club or not – claims to have 1.1 billion fans on the planet. With revenue of £488m in 2019-20, that’s just 45p per year, per fan.
“Clubs are thinking: ‘Can we ‘find another way of extracting money out of that huge fanbase?’ That’s where tokens come in.”
When AC Milan launched a token in early 2021, it raised $6m (£4.4m) in under an hour, or about 12% of the value of the club’s record signing, Leonardo Bonucci. Paris Saint-Germain’s token, the most valuable, has a market value of $45m.
In the murky and unregulated world of crypto though, it’s hard to know how much clubs are actually making. Socios said last year it had sold $300m worth of fan tokens but would not say how much of that went to the clubs with which it partnered.
Other platforms, such as Binance, are also moving into the fan token market, indicating there is room for growth, particularly given that only a few dozen clubs have entered the market in any meaningful way.
Pedro Herrera, senior blockchain analyst at DappRadar, a marketplace for blockchain-related apps, said that most fans buy tokens for the associated perks, such as votes on small decisions about which song to play over the stadium tannoy after a goal, or entry into a draw to win a signed shirt.
“It’s a win for the fan because they feel more involved; it’s a win for the team because it’s adding a layer of monetisation; and it’s a win for the [crypto] industry because you attract the masses and it’s one step closer to mass adoption.”
Maguire isn’t against crypto but adds a more sceptical tone: “Lots of fans love crypto and in its purest form it’s great. Banks have been overcharging people for years in terms of transaction fees and if crypto can reduce those fees that’s fantastic.”
“The problem is when unscrupulous traders, particularly via social media, seek to exploit fans who think a token is a serious investment product, rather than a glorified collectible.
“It’s magic beans. So long as it’s sold as a digital Panini card, it’s OK. But when it’s being seen as a form of investment, it’s moving into uncomfortable territory.”
“It’s unregulated, it’s volatile and it’s subject to manipulation by people who own large amounts of the asset.”
Fan tokens, though, are a mere paragraph in football’s rapidly unfolding crypto saga.
In 2021, crypto sponsors piled into football and were welcomed with open arms by cash-hungry clubs, leagues and players.
Exchange app Crypto.com sponsors Italy’s Serie A, one of the world’s most glamorous leagues, while Socios is Internazionale’s shirt sponsor. EToro, a trading platform that facilitates investment in multiple cryptocurrencies, has deals with more than half of the clubs in the Premier League.
Southampton players are understood to have been offered the option to be paid bonuses in bitcoin, as part of a £7.5m-a-year deal with Coingaming Group. And in January 2021, striker David Barral made history when he became the first player in a major league to be signed with bitcoin, albeit in Spain’s third tier with Internacional de Madrid.
This should come as no surprise given the reach that big-name stars have via social media and the money they can make from promotions. Other partnerships are perhaps more unexpected. Visitors to the Twitter profile of former Republic of Ireland international Tony Cascarino might have been wrongfooted by the former striker’s sudden change of pace midway through 2021. One moment he was musing on the latest developments in the Premier League, the next he was evangelising about blockchain bank Babb (no relation to former Ireland teammate Phil) and musing that the “crypto market is on fire”.
Even in its infancy, the reputational risks of this new commercial pact between crypto and football have become all too clear. Last year, Manchester City’s deal with a mysterious firm called 3Key Technologies fell apart in a matter of days as it emerged that nobody seemed to know anything about the company or its executives.
In December, Arsenal were rapped on the knuckles by the Advertising Standards Authority (ASA), which banned a club promotion that it said was exploiting fans’ “inexperience or credulity, trivialising investment in crypto assets, misleading consumers over the risk of investment and not making it clear the ‘token’ was a crypto asset”.
“For those in sport looking for sponsorship, it’s a whole new market of opportunity but it’s a bit of a landmine you’re dealing with,” said Bill Esdaile, managing director of sports marketing agency Square in the Air.
“My gut feeling is that such a small percentage of people understand how [crypto] works that too many decisions are made on trust, thinking that if [crypto firms] say they’ve got the money, they do.”
The amounts on offer appear to be going up.
Premier League strugglers Watford have perhaps the country’s biggest crypto deal, a front-of-shirt sponsorship from Stake.com. The site offers crypto gambling, which isn’t legal in the UK but may appeal to the league’s hundreds of millions of viewers around the world.
The arrangement even means that Watford players’ shirt sleeves bear the logo of Dogecoin, a “joke” currency whose value swings around wildly, often in response to tweets by Tesla multibillionaire Elon Musk.
Kieran Maguire estimates that the shirt deal could be worth up to £8m, based on the typical value of such partnerships, while an insider at Watford told MSN in August that the Dogecoin sleeve display added £700,000 into the mix.
Sums like these will become increasingly difficult for clubs to ignore, he thinks, particularly if the government goes ahead with a root-and-branch overhaul of gambling regulation that could see football lose the cash cow of shirt deals with betting firms.
“They [clubs] see the token market as slightly to one side, it won’t get picked up by the gambling review and it will help fill the gap,” says Maguire.
“Those deals of £5m to £8m could be replaced by NFT advertising and by crypto.”
In a recent paper, psychology researcher and gambling expert Dr Phil Newall warned that football sponsorship may be about to swap one risky product for another.
Researchers at Facebook parent’s Meta have trained a single AI model capable of processing speech, images, and text in the hope that these so-called multi-modal systems will power the company’s augmented reality and metaverse products.
The model, known as data2vec, can perform different tasks. Given an audio snippet, it can recognize speech. If it’s fed an image, it can classify objects. And when faced with text, it can check the grammar or analyse the writing’s tone and emotions.
AI algorithms are typically trained on one type of data, though data2vec is trained on three different modalities. It still, however, processes each form, whether its speech, images, and text, separately.
Meta believes these multi-modal models will help computers be more adaptable to blend physical and digital environments into one. “People experience the world through a combination of sight, sound and words, and systems like this could one day understand the world the way we do,” Meta CEO Mark Zuckerberg said in a statement to El Reg.
“This will all eventually get built into AR glasses with an AI assistant so, for example, it could help you cook dinner, noticing if you miss an ingredient, prompting you to turn down the heat, or more complex tasks.”
Data2vec is a transformer-based neural network and uses self-supervised learning to learn common patterns in audio, computer vision, and natural language processing. The model learns to operate with different types of data by learning how to predict how the representation of data it’s given; it knows it has to guess the next group of pixels when given an image, or the next speech utterance in audio, or fill in the words in a sentence.
The researchers used a mix of 16 Nvidia V100 and A100 GPUs to train data2vec on 960 hours of speech audio, millions of words from books and Wikipedia pages, and images from ImageNet-1K.
“We train separate models for each modality but the process through which the models learn is identical,” Alexei Baevski, a research engineer at Meta AI told The Register.
“We hope that it will enable future work to build high performing self-supervised models that combine modalities and are more effective than specialized models. Different modalities can add additional information to the same piece of content – for example body language from video, prosodic information from audio, and text can combine into a richer representation of a dialog. The algorithms that currently try to combine multi-modal information exist but they do not yet perform well enough to replace specialized algorithms and we hope our work will help change that.”
Baevski said in the future multi-modal systems could incorporate a larger range of data to model concepts like smell, 3D objects, or videos. He referred back to the idea of AR glasses helping wearers cook.
“Imagine having a model that has been trained on recordings of thousands of hours of cooking activity from various restaurants and chefs. Then, when you are cooking in a kitchen wearing your AR glasses that have access to this model, it’s able to overlay visual cues for what you need to do next, point out potential mistakes, or explain how adding a particular ingredient will affect the taste of your dish,” he told us.
Previous research on multi-modal systems have shown they can be prone to easy adversarial attacks. OpenAI’s CLIP model, for example, trained on images and text will identify an image of an apple incorrectly as an iPod if the word “iPod” are in the picture. It’s unclear, however, if data2vec suffers from similar weaknesses.
“We have not specifically analyzed how our models will react to adversarial examples but since our current models are trained separately for each modality, we believe that existing research on adversarial attack analysis for each modality would be applicable to our work as well,” Baevski said.
“In the future, we hope to use our work to enable high performance algorithms that combine modalities in one model and we plan to study how susceptible they are to adversarial attacks.”
When the researchers tested data2vec, it outperformed some top models that had been trained on a specific data type only on different types of tasks. The preliminary results are described in a paper [PDF], and the code has been published on GitHub.
“Data2vec demonstrates that the same self-supervised algorithm can work well in different modalities — and often better than the best existing algorithms,” the researchers explained in a blog post this week.
“This paves the way for more general self-supervised learning and brings us closer to a world where AI might use videos, articles, and audio recordings to learn about complicated subjects, such as the game of soccer or different ways to bake bread. We also hope data2vec will bring us closer to a world where computers need very little labeled data in order to accomplish tasks.” ®