British regulators have launched a preliminary investigation into Microsoft’s recent hiring spree from AI startup Inflection AI and its entry into associated arrangements with Inflection AI, due to concerns that this could hinder competition in the burgeoning AI market. Mustafa Suleyman, Inflection AI’s co-founder and CEO, along with several top engineers and researchers, joined Microsoft earlier this year. Suleyman, a co-founder of the AI research lab DeepMind, is a prominent figure in the AI industry.
The UK’s Competition and Markets Authority (CMA) is scrutinising whether these hirings might lead to a significant reduction in competition within the UK’s AI sector, potentially breaching antitrust regulations. Microsoft, however, maintains that the recruitment of talent fosters competition and should not be regarded as a merger. The company has pledged to cooperate with the CMA’s inquiry.
The CMA has a deadline of 11 September to decide whether to approve the hirings or escalate the investigation. The authority has the power to reverse deals or impose conditions to address any competition concerns. This investigation highlights the growing regulatory scrutiny over how major tech companies are acquiring talent and technology from innovative AI startups.
Across the Atlantic, US senators have urged antitrust enforcers to investigate Amazon’s deal with AI startup Adept. The senators noted similarities to the Microsoft-Inflection case, emphasizing concerns over the potential elimination of major competitors in the AI market. These developments reflect a broader regulatory focus on maintaining competitive balance in the rapidly evolving AI industry.
The Federal Communications Commission (FCC) has unveiled a proposal to tighten regulations on AI-generated robocalls. Chairwoman Jessica Rosenworcel seeks public input on defining AI-generated calls, mandating disclosure by those using such technology, and supporting technologies that alert consumers to unlawful AI robocalls. The proposal aims to balance consumer protection with enabling beneficial uses of AI for people with disabilities.
State attorneys general would gain new powers to target businesses using AI-generated robocalls under the proposal, which the FCC will vote on in August 2024. In February 2024, the FCC classified calls with AI-generated voices as ‘artificial’ under the Telephone Consumer Protection Act (TCPA), making robocalls using voice cloning technology illegal. The agency has highlighted the risk of misinformation through AI calls imitating celebrities, politicians, and family members.
Rosenworcel’s initiative follows a series of FCC actions against AI-generated robocalls, including significant fines and mandates for service providers to combat illegal robocalls. Recently, the FCC fined a service provider and political operative $8 million for a campaign using voice-cloned calls of President Joe Biden to mislead voters before the New Hampshire primary.
Microsoft researchers have unveiled SpreadsheetLLM, a new method designed to help AI better understand and process spreadsheets. Combining encoding and compression techniques, SpreadsheetLLM aims to enhance AI performance in handling expansive grids and various cell formats, which are typically challenging for language models.
Their data revealed that using this method, the GPT-4 model saw a 27% improvement in spreadsheet table detection and a 26% boost in performance for in-context learning. Additionally, the method led to up to 96% cost reductions when using GPT-4 and GPT-3.5-turbo models.
Central to this innovation is Microsoft’s SheetCompressor, which includes modules that make spreadsheets more legible for AI, bypass empty cells, and help AI understand the meaning of numbers.
Future integration of SpreadsheetLLM into Microsoft Copilot for 365 could simplify data analysis, allowing users to upload entire spreadsheets and ask AI chatbot questions in plain language for data summaries or analysis. Although the method currently can’t handle spreadsheet formatting details, its implementation could significantly ease user data management tasks.
Samsung plans to integrate generative AI (GenAI) into over 200 million smartphones by the end of the year, coinciding with the launch of its latest foldable phones and wearable devices. The company’s head of mobile business, Tim Roh, highlighted the use of Samsung’s Galaxy AI, which offers various productivity and creative features.
The tech giant is adopting a hybrid strategy, blending cloud computing with on-device processing to make this advanced technology accessible to more users. This approach requires sophisticated hardware, prompting Samsung to develop lightweight AI algorithms, a field in which they already lead the competition. Roh emphasised the necessity of optimising hardware to enhance AI capabilities while managing cloud costs.
Samsung’s R&D centres in Noida and Bengaluru are crucial in developing these AI features. Roh acknowledged the importance of the Indian market and Samsung’s collaboration with local partners to strengthen infrastructure and supply chains. While AI service monetisation is still under review, Samsung remains committed to enhancing user experiences and maintaining a competitive edge.
AI software designed to create multilingual tutorial videos for foreign workers in Japan has been launched. Tokyo-based Studist Corp developed ‘Teachme AI’ to help companies produce instructional videos quickly and efficiently.
Teachme AI can translate text into 20 different languages, including Thai, Vietnamese, Indonesian, and Bengali. This innovation aims to support businesses as the number of foreign workers in Japan rises, addressing labour shortages and an ageing population.
The software significantly reduces editing times, automatically dividing footage into chapters with subtitles. During a demonstration, a 30-minute video with Thai explanations was created in just 15 minutes, impressing users with its efficiency.
Google’s Gemini AI has been discovered scanning PDF files on Google Drive without user consent, sparking concerns over AI safety and privacy. Senior Advisor Kevin Bankston revealed that the AI generated a summary of a private tax return without permission, raising significant privacy issues.
Bankston shared his struggles to disable the feature, which continued to operate despite attempts to find the correct controls. The difficulty in managing Gemini’s integration in Google Drive has led to questions about Google’s handling of user data and privacy settings.
Google previously assured users that Workspace data would not be used to train AI or target ads. However, this incident has raised doubts about data hygiene and privacy.
Bankston’s experience suggests that prior participation in Google Workspace Labs might have influenced Gemini’s behaviour, highlighting the need for better user control and consent as AI technology advances.
A recent study has revealed that AI chatbots pose significant risks to children, who often view them as lifelike and trustworthy. Dr Nomisha Kurian from the University of Cambridge calls for urgent action to prioritise ‘child-safe AI’ in the development of these technologies.
Kurian’s research highlights incidents where AI chatbots provided harmful advice to children, such as Amazon’s Alexa instructing a child to touch a live electrical plug and Snapchat’s My AI giving tips on losing virginity.
These cases underscore the ’empathy gap’ in AI, where chatbots fail to respond appropriately to children’s unique needs and vulnerabilities.
The study proposes a 28-item framework to help developers create safer AI by working closely with educators and child safety experts. Kurian argues that AI has great potential if designed responsibly, but proactive measures are essential to protect young users.
Pittsburgh-based startup Skild AI has emerged from stealth mode with a $300 million funding round led by Jeff Bezos’ fund and Japan’s Softbank Group. Founded by Carnegie Mellon University professors Deepak Pathak and Abhinav Gupta, the company aims to build a ‘general-purpose brain’ for robots.
Skild AI’s AI model, trained on more data than its competitors, can be retrofitted to existing hardware. Pathak highlighted the model’s unparalleled capabilities across various robots and tasks, indicating its potential to transform automation in real-world environments.
With applications in household tasks and industrial work, Skild AI’s technology promises to reduce robot costs and address labour shortages. The company, now valued at $1.5 billion, is hiring engineers from top tech firms and Carnegie Mellon University.
According to a Reuters report, the fairly new OpenAI project, Strawberry, is set to create giant waves in the research industry. The project, which some claim could be a renamed version of the company’s project Q* from last year, has been tagged as potentially having capabilities to navigate the net to conduct deep research.
The company’s representative confirmed to the news agency that the reasoning ability of their models will invariably improve with time. Just last Tuesday, employees of OpenAI were treated to a demo of a model with human-like reasoning capabilities. The meeting came on the heels of the negative commentary the company has faced for placing a gag order on employees for publicly exposing the dangers its innovations can potentially pose to humanity.
Earlier in July, employees sent a seven-page letter to the US Security Exchange Commission (SEC) chair, Gary Gensler, detailing what they deem as risks OpenAI’s projects can pose to humans. The letter was tinged with urgency as the agency was advised to take swift and aggressive action against the company for violating current regulations.
Taiwan Semiconductor Manufacturing Co (TSMC), the leading producer of advanced chips for AI applications, is set to report a 30% rise in second-quarter profit on Thursday, driven by soaring demand. The world’s largest contract chipmaker, which counts Apple and Nvidia among its clients, has seen its stock—and the broader Taiwan market—reach record highs. Last week, TSMC’s American Depositary Receipts surpassed a trillion-dollar market value.
For the quarter ending 30 June, TSMC is expected to announce a net profit of T$236.1 billion ($7.25 billion), a significant increase from the T$181.8 billion reported in the same period last year. The company recently reported a substantial rise in second-quarter revenue, exceeding market expectations. Analysts like Li Fang-kuo from President Capital Management Co. anticipate a strong third-quarter outlook for all of TSMC’s products.
TSMC is heavily investing in expanding its production capacity, including spending $65 billion on three new plants in Arizona. However, the majority of its manufacturing will remain in Taiwan. The company is expected to maintain its capital expenditure guidance for this year at $28 billion to $32 billion, with 70% to 80% allocated to advanced technologies. KGI Securities’ Chu Yen-min suggests TSMC raise its capital spending due to favourable market conditions, which could further boost its stock price and support the broader market.
Why does this matter?
TSMC, often called the ‘sacred mountain protecting the country’ for its crucial role in Taiwan’s economy, remains a dominant player in the semiconductor industry despite challenges from Intel and Samsung. The AI boom has significantly increased TSMC’s stock price, which has surged 75% this year, outpacing the 33% gain in the broader market.
