Teachers have long worried about ChatGPT enabling students to cheat, with its ability to produce essays and solve problems quickly. According to Chegg, 40% of undergraduates worldwide have used generative AI, with many relying on it daily. However, education experts at the Fortune Brainstorm AI Singapore conference believe that generative AI can now be a valuable tool for educators.
Initially, educators feared widespread plagiarism with the introduction of ChatGPT, said Sonita Jeyapathy from the National University of Singapore. Yet, this concern has shifted as teachers began to see the potential of AI in enhancing their teaching methods. Khairul Anwar, founder of Malaysian edtech startup Pandai, noted that teachers seek AI assistance for lesson planning, student motivation, and professional development.
Pandai’s chatbot, designed to assist with homework, exemplifies this shift. Instead of providing direct answers, it guides students through problems step-by-step, encouraging more profound understanding. Tim Baldwin of the Mohamed bin Zayed University of Artificial Intelligence highlighted further AI advancements, such as personalised tutoring tailored to individual learning styles, making education more accessible.
Panelists acknowledged that AI-enabled cheating isn’t new and emphasised the importance of student motivation and institutional values in deterring dishonesty. Anwar suggested that framing education as a transactional path to material success can encourage cheating, advocating instead for highlighting the intrinsic value of learning.
Meta has informed investors that its heavy investments in generative AI won’t yield revenue this year but are expected to pay off in the long run. Founder Mark Zuckerberg explained that Meta is preparing for future computing needs, particularly its next AI model, Llama 4, which will require significantly more computing power than its predecessor. Despite the high costs, with capital expenditures projected to reach up to $40 billion this year, Meta aims to position Llama 4 as the industry’s most advanced model by next year.
CFO Susan Li acknowledged during the earnings call that generative AI would not contribute to this year’s revenue. However, she emphasised that Meta’s AI infrastructure is designed to be flexible and multifunctional, capable of supporting various AI needs, including content ranking and recommendations. Meta’s existing AI tools have already enhanced user engagement on platforms like Facebook and Instagram, and new video recommendation tools have further boosted interaction with features like Facebook Reels.
Zuckerberg also discussed AI’s potential to transform Meta’s advertising business. He predicted that AI would eventually handle tasks like writing personalised ad copy and managing campaign objectives, simplifying the process for advertisers. This vision aligns with Meta’s broader strategy of leveraging AI to improve user experience and business outcomes.
Why does this matter?
Despite significant losses in its Reality Labs division, which focuses on metaverse-related technology, Meta reported solid financial performance. The company showed robust growth with quarterly revenues of $39 billion and net income of $13.5 billion. Additionally, Meta’s daily user base reached 3.2 billion, and its new social media app, Threads, is nearing 200 million monthly active users, contributing to investor confidence and a rise in Meta’s share price.
Telford and Wrekin Council, the local authority of Telford and Wrekin in the ceremonial county of Shropshire, England, have introduced an upgraded AI phone assistant named Ask Tom, designed to improve customer service by handling queries without the need for human staff.
Initially launched in 2021 as a digital assistant, Ask Tom now has the capability to answer, listen, speak, and send information directly to customers’ devices. The enhancement aims to free up customer service staff to focus on more complex issues, although human advisors will still be available if needed.
The council highlighted that when Ask Tom was first introduced, it significantly reduced customer waiting times by 50%. With the new upgrade, the council expects the system to assist even more people by addressing multiple questions in a single call and preventing the need to contact different numbers for specific information.
Councillor Zona Hannington emphasised that the automated assistant is built on a well-tested, extensive knowledge base of council information. The council can allocate its staff to more critical tasks by automating simple queries, thus improving efficiency and effectiveness. Residents will benefit from 24-hour access to a broader range of council services, enhancing the overall service experience.
Meta’s AI ambitions have received a significant boost from WhatsApp’s 500 million users in India. During Meta’s second-quarter earnings call, CFO Susan Li highlighted that India has become the largest market for Meta AI usage, a remarkable achievement given the product’s recent launch. Users in India have generated billions of queries, showing high engagement and retention, particularly on WhatsApp.
Meta AI was initially launched in the US and expanded to India and over a dozen other countries with the upgraded Llama 3 model. Despite initial cultural challenges, such as generating images and jokes that were culturally insensitive, Meta has adjusted its algorithm to suit the Indian market better. Additionally, the AI now supports multiple languages beyond English, including Hindi and Spanish, broadening its accessibility.
CEO Mark Zuckerberg emphasised the increasing computational demands for future AI models, stating that training the upcoming Llama 4 model would require nearly ten times the computing power used for Llama 3 and so demonstrating Meta’s commitment to advancing its AI capabilities.
Apart from AI, Meta’s social networks also show promise. The Threads platform is nearing 200 million users, and Facebook usage among young adults in the US is rising, indicating continued growth and engagement across Meta’s suite of products.
OnlyFans, a platform known for offering subscribers ‘authentic relationships’ with content creators, faces scrutiny over the use of AI chatbots impersonating performers. Some management agencies employ AI software to sext with subscribers, bypassing the need for human interaction. NEO Agency, for example, uses a chatbot called FlirtFlow to create what it claims are ‘genuine and meaningful’ connections, although OnlyFans’ terms of service prohibit such use of AI.
Despite these rules, chatbots are prevalent. NEO Agency manages about 70 creators, with half using FlirtFlow. The AI engages subscribers in small talk to gather personal information, aiming to extract more money. While effective for high-traffic accounts, human chatters are still preferred for more personalised interactions, especially in niche erotic categories.
Similarly, Australian company Botly offers software that generates responses for OnlyFans messages, which a human can then send. Botly claims its technology is used in over 100,000 chats per month. Such practices raise concerns about transparency and authenticity on platforms that promise direct interactions with creators.
The issue coincides with broader discussions on online safety. The UK recently amended its Online Safety Bill to combat deepfakes and revenge porn, highlighting the rising threat of deceptive digital practices. Meanwhile, other platforms like X (formerly Twitter) have officially allowed adult content, increasing the complexity of managing online safety and authenticity.
A formal complaint has been filed with the Agency for Access to Public Information (AAIP) of Argentina against Meta, the parent company of Facebook, WhatsApp and Instagram. The case is in line with the international context of increasing scrutiny on the data protection practices of large technology companies.
The presentation was made by lawyers specialising in personal data protection, Facundo Malaureille and Daniel Monastersky, directors of the Diploma in Data Governance at the CEMA University. The complaint signals the company’s use of personal data for AI training.
The presentation consists of 22 points and requests that Meta Argentina explain its practices for collecting and using personal data for AI training. The AAIP will evaluate and respond to this presentation as the enforcement authority of the Personal Data Protection Law of Argentina (Law 25,326).
The country’s technological and legal community is closely watching the development of this case, given that the outcome of this complaint could impact innovation in AI and the protection of personal data in Argentina in the coming years.
AMD has increased its 2024 forecast for AI chip sales by $500 million, anticipating continued tight supplies through 2025. The announcement sent shares of the California-based company up by 7.5% in extended trading. AMD’s AI chips, mainly purchased by cloud computing giants, are considered strong competitors to Nvidia’s dominant market presence. Following AMD’s report, Nvidia shares also rose by 4.7%.
CEO Lisa Su revealed that AMD’s AI chip revenue forecast 2024 is now $4.5 billion, up from the previous $4 billion target. In the second quarter, AMD’s data centre revenue, which includes AI chip sales, surged by 115% to $2.8 billion, surpassing estimates. For the first time, quarterly AI chip revenue exceeded $1 billion, highlighting the growing demand for these products.
Despite the significant investment in AI technology, substantial returns have yet to materialise. The outcome was evident as shares of Microsoft fell 6% due to missed growth targets in its cloud-computing services, suggesting that benefits from AI investments may take longer to materialise.
AMD’s revenue in the second quarter rose 9% to $5.8 billion, exceeding expectations. The company also forecasts third-quarter revenue of $6.7 billion, with a gross margin of about 53.5%. Additionally, AMD’s PC chip business is experiencing a recovery, reporting second-quarter revenue of $1.5 billion, driven by hopes that new AI features will boost consumer demand.
Microsoft plans to increase its spending on AI infrastructure this fiscal year despite slower growth in its cloud business. This announcement led to a 4% drop in its share price after an initial 7% decline. The tech giant, along with others like Google, is investing heavily in data centres to leverage the AI boom, with Microsoft’s capital spending rising 77.6% to $19 billion in its fiscal fourth quarter, primarily for cloud and AI-related expenses.
Despite these investments, investors were disappointed with the slower growth of Microsoft’s Azure cloud service. The company forecasted a 28% to 29% growth for Azure in the upcoming quarter, slightly below market expectations, which followed a 29% increase in the previous quarter, but it also fell short of estimates, indicating a slowdown from earlier months.
CEO Satya Nadella highlighted that AI services are becoming a significant part of Azure’s revenue growth, with over 60,000 customers using Azure AI, a nearly 60% increase from the previous year. Microsoft has integrated AI across its products, including its search engine Bing and productivity tools like Word, driven by its substantial investment in OpenAI.
Microsoft’s total revenue rose 15% to $64.7 billion in the fourth quarter, exceeding analyst expectations. The company also grew in its personal computing business, benefiting from stabilising PC sales. However, revenue from its Intelligent Cloud unit, which includes Azure, missed analyst estimates, rising 19% to $28.5 billion.
Brazil has announced a 23 billion reais ($4.07 billion) investment plan for AI development. The initiative aims to foster sustainable and socially-oriented technologies within the nation, enhancing its technological autonomy and competitiveness in the global AI market.
The investment plan includes immediate impact initiatives targeting key sectors such as public health, agriculture, environment, business, and education. These initiatives focus on developing AI systems to streamline customer service and operational procedures.
A significant portion of the funds, nearly 14 billion reais, will be allocated to business innovation projects over the next four years. More than 5 billion reais will be invested in AI infrastructure and development, with the remaining resources dedicated to training, public service improvements, and AI regulation support.
President Luiz Inácio Lula da Silva emphasised the importance of Brazil developing its own AI technologies rather than relying on imports. He highlighted the potential of AI to generate income and employment within the country.
Apple Inc has joined US President Joe Biden’s voluntary commitments to govern artificial intelligence, aimed at preventing the misuse of AI technology. The White House announced on Friday that Apple is now part of a group of 15 firms that have committed to ensuring AI’s power is not used for harmful purposes. The original commitments, introduced in July 2023, were initially signed by companies such as Google and Microsoft’s partner OpenAI.
In September, additional firms including Adobe, IBM, and Nvidia also pledged their support. This initiative is part of a broader effort by the Biden administration to promote responsible AI innovation by assembling an AI expert team, urging tech CEOs to adopt measures that prevent AI from being used destructively.
Apple’s participation comes amid its own challenges with AI, as the company recently delayed AI features for iOS and iPadOS. This commitment underscores the importance of a unified approach among major tech companies to address the ethical and safety concerns surrounding AI.
