ai

Figure AI tackles workplace safety for robots

Figure AI has announced the creation of the Centre for the Advancement of Humanoid Safety, a new initiative aimed at ensuring humanoid robots can operate safely in workplaces. Led by former Amazon Robotics safety engineer Rob Gruendel, the centre will focus on testing AI-controlled robots for stability, human detection, and navigation to minimise accidents.

The rise of humanoid robots in warehouses and factories has sparked concerns about their potential risks. Unlike traditional industrial robots, which were confined to cages, these machines move freely among workers, raising safety questions. Existing solutions, such as Amazon’s wearable safety vest and Veo Robotics’ vision-based systems, have helped, but regulation remains largely absent.

Figure AI plans to release regular safety reports detailing its progress, testing methods, and solutions for potential hazards. As companies push to integrate humanoid robots into daily operations, and eventually, into homes, the need for clear safety standards is becoming increasingly urgent.

AI developers gain more flexibility on hugging face

Hugging Face has introduced Inference Providers, a new feature that allows developers to run AI models on third-party cloud services. Partnering with companies like SambaNova, Fal, Replicate, and Together AI, the platform now offers users the flexibility to deploy models on different infrastructures directly from their project pages.

Previously, Hugging Face primarily focused on its in-house AI hosting solutions, but the company is shifting towards a more collaborative approach. By integrating with external serverless providers, developers can now scale their models without managing hardware, making deployment easier and more cost-efficient. Users will pay standard provider rates, and Hugging Face Pro subscribers will receive additional free credits.

Since its founding in 2016, Hugging Face has grown into a leading AI model hub, backed by major investors like Google, Amazon, and Nvidia. With its latest move, the company continues to expand its ecosystem, making AI more accessible for developers worldwide.

Italy blocks DeepSeek chatbot over privacy concerns

Italy’s data protection authority, the Garante, has ordered the Chinese AI startup DeepSeek to block its chatbot in the country, citing insufficient responses to queries about its privacy policy. The watchdog had requested detailed information on data collection practices, sources, purposes, and storage, particularly concerning whether user data is stored in China. DeepSeek’s failure to adequately address these concerns prompted the Garante to impose an immediate ban and launch an investigation.

DeepSeek had removed its AI assistant from Italian app stores earlier this week but claimed it was not subject to local regulation. Agostino Ghiglia, a member of the Garante’s board, stated that the company’s stance worsened its position. Italian users who had already downloaded the app still reported access to the chatbot, while the web version remains operational. The Garante emphasised that European citizens must have clear consent and data protection guarantees, especially regarding servers located in China.

The Garante’s action highlights growing scrutiny of AI platforms in Europe, with data regulators in Ireland and France also questioning DeepSeek’s privacy practices. Italy‘s proactive approach has drawn attention; the country temporarily banned ChatGPT in 2023 over similar concerns. DeepSeek has positioned its AI as a cost-effective alternative to US models, surpassing ChatGPT as the top-rated app on Apple’s US App Store. However, its refusal to cooperate with European regulators may jeopardise its expansion.

Microsoft faces pressure ahead of AI growth forecast

Microsoft’s upcoming quarterly forecast will reveal whether its significant investments in AI, including its partnership with OpenAI, drive growth in its key Azure cloud business. Despite earlier optimism, Azure’s growth has slowed for two consecutive quarters, and investors are anxious about Microsoft’s ability to monetise AI. The company has committed about $80 billion in capital spending this year, but doubts linger over the effectiveness of its strategy, especially after a sharp drop in stock price following the launch of a competitive AI model by Chinese startup DeepSeek.

Azure, which contributes around a third of Microsoft’s revenue, is expected to show 31.8% growth in the second quarter, a slight slowdown from the previous quarter. Microsoft’s relationship with OpenAI remains a key growth driver, with Azure set to handle much of OpenAI’s cloud traffic. However, investor sentiment has soured, with growing concerns about AI monetisation, margins, and capital expenditure. Microsoft also faces the impact of a stronger dollar, which could hurt its international earnings.

In addition to Azure, Microsoft is banking on the success of its Microsoft 365 Copilot AI assistant, but adoption has been slower than anticipated. To stimulate demand, the company has adjusted its pricing, adding AI features to lower-tier Microsoft 365 plans. While the Copilot’s potential remains high, analysts project a modest penetration rate of 10%, suggesting it could add significant revenue in the coming years. Despite these challenges, Microsoft’s productivity division, which includes 365 Copilot and LinkedIn, is expected to see continued growth.

Overall, Microsoft is forecasted to report slower growth for the second quarter, with revenue expected to rise by 10.9% compared to 16% in the first quarter. Net profit is also projected to increase at a slower pace, raising questions about whether the company’s AI investments will pay off as anticipated.

India’s copyright lawsuit targets OpenAI and AI use

Microsoft-backed OpenAI is seeking to prevent some of India’s largest media organisations, including those linked to Gautam Adani and Mukesh Ambani, from joining a copyright lawsuit. The case, initiated by news agency ANI last year, involves claims that AI systems like ChatGPT use copyrighted material without permission, sparking a wider debate over AI and intellectual property in the country. India ranks as OpenAI’s second-largest market by user numbers, following the US.

OpenAI has argued its AI services rely only on publicly available data and adhere to fair use principles. During Tuesday’s hearing, OpenAI’s lawyer opposed bids by additional media organisations to join the case, stating he would submit formal objections in writing. The company has also challenged the court’s jurisdiction, asserting that its servers are located outside India. The case is scheduled to continue in February.

The Federation of Indian Publishers has accused ChatGPT of harming their business by summarising books from unlicensed online sources. OpenAI denies these claims, maintaining its tools do not infringe copyright. Prominent digital media groups, including the Indian Express and Hindustan Times, allege ChatGPT scrapes and reproduces their content, prompting their involvement in the lawsuit.

Tensions escalated over media coverage of the case, with OpenAI objecting to reports based on non-public court filings. Lawyers representing media groups called such claims unfounded. The lawsuit is poised to shape the future of AI and copyright law in India, as courts worldwide grapple with similar challenges.

Will DeepSeek rise strenghten open-source AI in the United States?

DeepSeek, a Chinese AI company, is rapidly transforming the global artificial intelligence landscape. The models’ standout feature is their open-source nature, which allows developers worldwide to access, modify, and build upon them. This contrasts with the closed-source strategies of most American tech firms, barring Meta, fostering an environment of transparency and innovation that challenges traditional industry norms.

This shift marks a significant realignment in technological power dynamics as DeepSeek elevates China alongside open-source models to compete with the traditionally dominant American, closed-source approaches. Such advancements have impacted stock markets, causing US tech stocks to drop and signalling a shift in competitive balance between the United States and China, as well as open- versus closed-source models.

DeepSeek’s methodologies question traditional AI training practices, expediting the race towards more efficient, lower-cost processes without sacrificing performance. They have advanced public understanding of model training through self-reflection and reasoning, democratising AI developments previously shielded by companies like OpenAI.

DeepSeek’s approach highlights the potential for cost-effectiveness in AI training, reducing reliance on extensive pretraining by utilizing reinforcement learning on open models, such as Meta’s Llama 3. This facilitates access for less-resourced research labs, enabling wider contributions to AI advancements.

The ongoing competitive landscape suggests a balance could be struck between open- and closed-source offerings, placing increasing pressure on American tech firms to reconsider their strategies. In response, the US government has initiated the Stargate Project, pledging $500 billion over the next four years to fortify AI infrastructure, underscoring the need to maintain America’s competitive edge.

In conclusion, DeepSeek’s rise endorses a trend towards open-source AI, challenging the entrenched power structures within the tech industry. By demonstrating efficient methodologies and promoting a collaborative developmental environment, DeepSeek not only illuminates open-source models’ potential to spearhead innovation but also highlights the shifting global landscape of AI research. As this competition unfolds, the dominance of traditional American firms may be reassessed, with open-source models presenting a promising and strategic path for the future of global AI ecosystems.

Australian shares hit by DeepSeek’s rise in AI

The launch of DeepSeek’s cost-efficient AI model has sent shockwaves through Australian tech markets, with shares in AI-related companies experiencing steep declines. Investors are increasingly worried that the Chinese startup’s affordable technology could undermine the dominance of established players in the sector.

Among the biggest losers were AI software firm Appen, which saw its stock drop by 3.3%, and chipmaker Brainchip, which lost 10.3%. The technology sub-index fell by 1%, with major data centre operators also taking a hit. Analysts expressed concerns that DeepSeek’s success might reduce demand for AI infrastructure, which had driven heavy investments in Australian data centres.

DeepSeek’s AI assistant, launched last week, has already outpaced US competitor ChatGPT in downloads on Apple’s App Store. This rapid rise has sent ripples through the global tech sector, contributing to Nvidia’s record $592.7 billion market loss.

As Australian investors reassess their exposure to AI stocks, market strategists predict a shift towards safer sectors such as healthcare and consumer staples, after DeepSeek’s disruptive impact.

Nvidia sees record retail investment amid stock plunge

Retail investors made a record purchase of Nvidia shares on Monday, buying a net $562.2 million worth of stock, following a sharp 17% drop in its market value. The decline came after concerns arose over a low-cost AI model from Chinese startup DeepSeek, which contributed to Nvidia losing $593 billion in market value. According to Vanda Research, this marked the largest retail investment in Nvidia since data tracking began in 2014.

Nvidia has seen steady retail investment over the past few years, with approximately $7.3 billion in shares purchased last quarter. However, this was nearly half the amount recorded in the peak quarter of September 2024.

While global tech stocks showed some recovery on Tuesday, the sector remains under pressure as investors grapple with concerns over the high valuations and dominance of AI leaders like Nvidia, amid rising competition from new players like DeepSeek.

SAP sees stronger demand for AI services

SAP anticipates growing demand for its AI services as energy-efficient models become more accessible. The company’s CEO, Christian Klein, stated that advancements like China’s DeepSeek model signal a shift toward scalable AI infrastructure, which is crucial for running general AI systems effectively.

Klein emphasised the increasing competition in the AI landscape, suggesting that the market will feature multiple large language models rather than dominance by a single player. He noted these developments as positive for SAP’s position in the industry.

Despite recent volatility in technology stocks triggered by concerns over the profitability of AI investments, SAP remains optimistic. The potential of cheaper and more efficient AI systems could reinvigorate confidence in the sector, benefiting companies with robust AI offerings.

Italian data protection authority launches investigation into DeepSeek AI over privacy concerns

The Italian Data Protection Authority (IDPA) has commenced an investigation into DeepSeek, an AI firm, raising concerns about privacy and the handling of personal data. This initiative follows the IDPA’s pattern of scrutinising major AI platforms, as observed in past evaluations of ChatGPT.

The Authority has requested detailed information from Hangzhou DeepSeek Artificial Intelligence and Beijing DeepSeek Artificial Intelligence, the entities responsible for the DeepSeek chatbot service, in both its online and app-based forms. The central concern stems from the potential high risk to the personal data of millions of Italians, prompting the IDPA to question DeepSeek’s data collection sources, purposes, legal grounds, and storage locations, particularly focusing on any servers in China.

Further, the IDPA inquiry delves into the data utilised to train DeepSeek’s AI system. It seeks specific clarifications on the processing of personal data, especially if obtained through web scraping, aiming to assess how registered and non-registered users are informed about their data usage. This aspect underscores the priority of transparency and the necessity for obtaining informed consent. The investigation’s urgency is highlighted by the 20-day deadline given to DeepSeek to furnish the requisite information, characterising the IDPA’s approach as swift and firm given potential privacy risks within the EU.

This action by the IDPA coincides with ‘Data Protection Day’, emphasising the agency’s commitment to privacy issues and enhancing international awareness regarding data protection. It reflects a broader regulatory trend of increased scrutiny over AI technologies, ensuring stringent data protection standards are upheld. As AI advances, ensuring public trust and safeguarding personal information are deemed critical, and such regulatory measures are likely to become more prevalent.