Since his defeat against Google's DeepMind's AI program, AlphaGo, in 2016, Lee Saedol's life has fundamentally changed. The former 18-time world champion Go player, shocked by his loss against AI, went on to retire three years later. "I could no longer enjoy the game," he said, seeing the game he saw as an art form won by "an algorithm's ruthless efficiency." Yet the technology that bested him became an object of fascination for Lee.
After observing its breakthroughs in various fields, Lee now delivers lectures about AI, urging others to familiarize themselves with it. While he believes that AI may replace specific jobs, for him, the technology is bound to create some as well. Still, the former champion worries about the impact of AI on what humans value. "People used to be in awe of creativity, originality and innovation," says Lee. "But since AI came, a lot of that has disappeared." Still, Lee continues to be involved in the world of Go, having written several books on the subject and founded a Go academy for children.
Find more at the New York Times
The European Commission has rejected Meta's "pay or consent" subscription model, arguing that it is illegal under the bloc's new Digital Markets Act (DMA). The model, introduced in November 2023, asked Europeans to pay up to €12.99 ($14) per month for their privacy. Meta denied its subscription model broke the rules, stating that it follows the direction of the highest court in Europe and complies with the DMA.
The Commission's main concern is not that Meta is charging for an ad-free service, but that there should be a third option that may still contain ads but are just less targeted. Under the DMA, large tech platforms must ask users for consent if they want to share their personal data with other parts of their businesses. The Commission is particularly concerned about the competitive advantage Meta receives over its rivals by being able to combine the data from platforms like Instagram and its advertising business. If Meta cannot reach an agreement with regulators before March 2025, the Commission has the power to levy fines of up to 10 percent of the company’s global turnover.
Read more at Wired and the New York Times
Governments worldwide are increasingly relying on AI chatbots to automate their services and provide advice to citizens. The UK's Government Digital Service has tested a ChatGPT-based chatbot, GOV.UK Chat, which can answer questions about government services. While 70% of trial participants found the responses useful, there were instances of the system presenting incorrect information as factual. Likewise, Portugal has also launched a chatbot, the Justice Practical Guide, which answers basic questions about topics like marriage, divorce, and setting up a company, yet struggles with more complex queries, highlighting the need for further improvements in accuracy and reliability.
Experts caution against using chatbots as a replacement for human civil servants due to issues of accountability and reliability. Colin Van Noordt, a researcher on the use of AI in government, advises instead to view them as “an additional service, a quick way to find information.
Estonia, a leader in digitizing public services, also developed a suite of chatbots for state services called Bürokratt. But unlike the UK and Portugal's chatbots, Estonia's chatbots use Natural Language Processing (NLP), a technology that predates the latest wave of AI. While NLP models have limitations in imitating human speech and detecting nuances in language, they are less likely to provide incorrect or misleading answers.
More details at the BBC
Discover all the latest news, articles, webinar replays and fifty-five events in our monthly newsletter, Tea O'Clock.
%20(1).png)
.svg)
