AI Program Aims to Enhance Wikipedia’s Reliability, Okta’s Support System Breached and Customer Files Accessed, Microsoft CEO Nadella’s Compensation Decreases to $48.5M
These and more top tech stories on Hashtag Trending
I’m your host James Roy, filling in for Jim Love today.
Researchers have developed SIDE, an artificial intelligence program designed to improve the reliability of Wikipedia references. The AI’s primary functions are to verify the accuracy of a primary source and suggest new ones. However, SIDE operates under the assumption that a Wikipedia claim is genuine, meaning it can validate a source but cannot verify the actual claims made in an entry. In tests, the AI’s suggested citations were preferred over the original ones 70 per cent of the time. Notably, in almost half the cases, SIDE recommended a source already used by Wikipedia. Despite its potential, SIDE has limitations, such as only considering web page references. The researchers acknowledge that while AI can aid in fact-checking, biases and the nature of Wikipedia’s open editing could pose challenges.
Sources: Engadget
Okta, a prominent identity and authentication management provider, has confirmed that hackers accessed its customer support management system, viewing private customer data. The breach allowed the attackers to view files uploaded by certain Okta customers related to recent support cases. Okta’s Chief Security Officer, David Bradbury, stated that these files often included HTTP archive (HAR) files, which company support personnel use for troubleshooting, and can contain sensitive data like cookies and session tokens. Such data can be exploited by malicious actors to impersonate legitimate users. The breach was initially detected by security firm BeyondTrust, which observed suspicious activity on its in-house Okta administrator account. BeyondTrust alerted Okta but claimed they received no response for over two weeks. Okta has faced multiple security incidents in the past, including breaches in March and December 2022.
Sources: Arts Technica
Amazon-backed startup Anthropic, known for its Claude generative AI chatbot, is pioneering a novel approach to responsible AI by developing a constitution for its large language models based on public feedback. Recognizing the controversies surrounding AI control and influence, Anthropic polled 1,000 Americans to understand their desired values and guardrails for AI. The survey results yielded 75 principles, which were then compared to Anthropic’s existing 58 principles. Interestingly, only half of the survey’s principles overlapped with Anthropic’s, underscoring the tech industry’s potential disconnect from the broader public. When tested, both sets of principles produced similar accuracy results on the Claude model. However, the public’s “constitution” was found to be “less biased” across various social categories. This initiative emphasizes the importance of public involvement in shaping AI’s foundational values.
Sources: Axios
Amazon has updated its guidelines, granting managers the authority to terminate employees who refuse to comply with the company’s return-to-office mandate of being present at least three times a week. The details of this new policy were not fully accessible in the provided content, but it indicates a shift in Amazon’s stance towards remote work, and the importance the company places on in-office presence.
Sources: Business Insider
Satya Nadella, (chief executive officer) CEO of Microsoft, received a compensation of $48.5 million in 2023, marking the lowest amount in the past three years due to reduced stock awards and other incentives. This reduction comes after Nadella, who has amassed an estimated $1 billion since becoming CEO in 2014, informed full-time salaried employees of a pay freeze due to a revenue growth slowdown. The compensation breakdown includes a $2.5 million salary, $39.23 million in stock, $6.4 million in non-equity incentives, and $360k for other compensations. Despite this, Microsoft reported an 11 per cent revenue growth, reaching $211.9 billion. Over 95 per cent of Nadella’s annual target compensation is performance-based. The CEO’s pay ratio to the median employee’s compensation is 250 to 1, with the “median employee earning $193,770 annually.”
Sources: The Register
That’s the top tech news stories for today. For more fast reads on top stories, check us out at TechNewsDay.com or ITWorldCanada.com on the homepage.
Hashtag Trending goes to air 5 days a week with a special weekend interview show we call “the Weekend Edition.”
You can get us anywhere you get audio podcasts and there is a copy of the show notes at itworldcanada.com/podcasts
I’m your host, James Roy – have a Marvelous Monday.The post Hashtag Trending Oct.24 – Researchers have developed an AI program to improve Wikipedia’s reliability; Okta confirms hackers accessed customer data; Anthropic pioneers responsible AI through public feedback first appeared on IT World Canada.