Harvard University announced Thursday it's releasing a high-quality dataset of nearly one million public-domain books that could be used by anyone to train large language models and other AI tools. From a report: The dataset was created by Harvard's newly formed Institutional Data Initiative with funding from both Microsoft and OpenAI. It contains books scanned as part of the Google Books project that are no longer protected by copyright.

Around five times the size of the notorious Books3 dataset that was used to train AI models like Meta's Llama, the Institutional Data Initiative's database spans genres, decades, and languages, with classics from Shakespeare, Charles Dickens, and Dante included alongside obscure Czech math textbooks and Welsh pocket dictionaries. Greg Leppert, executive director of the Institutional Data Initiative, says the project is an attempt to "level the playing field" by giving the general public, including small players in the AI industry and individual researchers, access to the sort of highly-refined and curated content repositories that normally only established tech giants have the resources to assemble. "It's gone through rigorous review," he says.

Leppert believes the new public domain database could be used in conjunction with other licensed materials to build artificial intelligence models. "I think about it a bit like the way that Linux has become a foundational operating system for so much of the world," he says, noting that companies would still need to use additional training data to differentiate their models from those of their competitors.

The country's poorer communities often lack access to tech equipment. Teams that recruit low-income players are providing another path to economic mobility. From a report: On the outskirts of the most diverse cities in Brazil lie neighborhoods that climb steep hills and stretch for miles. These neighborhoods often have a precarious structure -- houses built side by side, with no apparent order, and only small corridors that are poorly lit. It is in these favelas that thousands of Brazil's youth dedicate hours and hours of their days to esports, with the dream of making it big in the industry. Projections point to a market that, in 2023, should surpass $1.5 billion, and in Brazil even traditional football teams such as Vasco da Gama and Flamengo have begun to assemble esports teams in games such as League of Legends and Pro Evolution Soccer. The top athletes win millions of dollars in prizes, while the average salary of a professional League of Legends player exceeds $400,000 per year.

Brazil is an extremely unequal country with an immense social abyss -- about 25 percent of the Brazilian population is considered poor, according to the Brazilian Institute of Geography and Statistics. Social inequality in Brazil, according to the Gini index (used by the World Bank to measure inequality among countries or groups of people), has increased in recent years. In regions like the northeast, almost half the population lives in extreme poverty on less than $1.90 a day. This inequality is also reflected in the country's esports industry. The basic items an esports athlete or streamer needs -- access to the internet and quality equipment -- is not always available for those living in the favelas. In an extremely competitive environment where fractions of a second can make all the difference and lead to a victory, slow internet and outdated equipment can be fatal to success. There are immense differences between those living in the favela and those in the "asphalt" -- that is, people living outside poor communities, who have access to better schools, health services, and greater purchasing power, and who often frown upon those from the favela.

In countries where websites, blogs and a free press are strictly limited, Minecraft "is still accessible by everyone," notes the official official web site for Reporters Without Borders (an international nonprofit defending freedom of information): Reporters Without Borders used this backdoor to build "The Uncensored Library": A library that is now accessible on an open server for Minecraft players around the globe. The library is filled with books, containing articles that were censored in their country of origin. These articles are now available again within Minecraft hidden from government surveillance technology inside a computer game. The books can be read by everyone on the server, but their content cannot be changed. The library is growing, with more and more books being added to overcome censorship.

On March 12 — the World Day Against Cyber Censorship — the Uncensored Library will open its doors, giving young people around the world access to independent information, through a medium they can playfully interact with. The campaign runs under the hashtag: #TruthFindsAWay...

Additional to banned articles from journalists, visitors of The Uncensored Library can find the Reporters Sans Frontières World Press Freedom Index and reports on the current press freedom situation of 180 countries in the world.
They spent three months working with the design studio BlockWorks to assemble 12.5 million blocks into a vast structure with a neoclassical architecture.

You can see it in a short film at uncensoredlibrary.com, or access it in Minecraft at visit.uncensoredlibrary.com

HughPickens.com writes: Joe Drape and Jacqueline Williams report at the NYT that a major scandal is erupting in the multibillion-dollar industry of fantasy sports, the online and unregulated business in which an estimated 57 million people participate where players assemble their fantasy teams with real athletes. Two major fantasy sports companies were forced to release statements defending their businesses' integrity after what amounted to allegations of insider trading — that employees were placing bets using information not generally available to the public. "It is absolutely akin to insider trading. It gives that person a distinct edge in a contest," says Daniel Wallach. "It could imperil this nascent industry unless real, immediate and meaningful safeguards are put in place."

In FanDuel's $5 million "NFL Sunday Million" contest this week, DraftKings employee Ethan Haskell placed second and won $350,000 with his lineup that had a mix of big-name players owned by a high number of users. Haskell had access to DraftKings ownership data meaning that he may have seen which NFL players had been selected by DraftKings users, and by how many users. In light of this scandal, DraftKings and FanDuel have, for now, banned their employees from playing on each other's sites. Many in the highly regulated casino industry insist daily fantasy sports leagues are gambling sites and shouldn't be treated any differently than traditional sports betting. This would mean a high amount of regulation. Industry analyst Chris Grove says this may be a watershed moment for a sector that may need the legislation it has resisted in order to prove its legitimacy. "You have information that is valuable and should be tightly restricted," says Grove. "There are people outside of the company that place value on that information. Is there any internal controls? Any audit process? The inability of the industry to produce a clear and compelling answer to these questions to anyone's satisfaction is why it needs to be regulated."

ClockEndGooner writes Professional club football in Europe, or soccer, as it's known here in the States, is perhaps the most expensive and costly professional team sport in the world. Yesterday, Spain's traditional powerhouse, Real Madrid, fielded a starting eleven roster that cost the club over $637 Million (£382 Million Pounds Sterling) to acquire and assemble over the past six seasons against rival club Sevilla in the UEFA Super Cup match played in Cardiff, Wales. With billions of dollars spent by the top teams in the world's most competitive leagues in Europe, and billions more at stake from TV royalties and commercial licensing rights, its crucial talent scouts, general managers or "gaffers", sporting directors and club owners and the rest of their back office staff do their homework before recruiting and signing new players. Prozone Sports Ltd. has turned to game publisher Sports Interative's popular Football Manager video game to include more player data and archived video footage of tens of thousands of players from across the world in its Prozone Recruiter application to help clubs make better and more informed decisions on player performances and strengths. Though not officially published, it is known that many of the top clubs in England, Spain, Germany, Italy, France, Holland and Russia rely on Prozone Recruiter.

Katamari. Since the launch of the original title, it's been a hat, a cake, and now it's another game. Even then, saying it is another game is a stretch. For the most part the aptly titled We Love Katamari is a mission pack for the original game with an expanded multiplayer component. I don't really care, and anyone who has had the chance to play the original isn't likely to care either. The game is pure PS2 gold, just as much fun as the first foray into star-making. We Love Katamari is pure fan service from the man who doesn't even like games that much anyway. Read on for my impressions of the sequel to the original star-creation-through-rolling simulator.

John R. Zedlewski has contributed an excellent feature entitled 'Free Software and the Innovators Dilema'. Talks a lot about how industries tend to shift, and what happens when a new low end/low cost technology wrecks the margins. Its worth a read. Check it out.