Online fashion retailers are increasingly using artificial intelligence to tackle the costly issue of clothing returns, with up to 30% of purchases being sent back due to sizing problems. A study by McKinsey estimates that each return costs between $21 and $46, significantly affecting profit margins. Many customers order multiple sizes and return those that don’t fit, creating logistical headaches for retailers.
To address this, companies are adopting AI-driven sizing tools. French start-up Fringuant, for instance, uses an algorithm that analyses a shopper’s height, weight, and a quick selfie to predict the best size. Zalando, a German e-commerce giant, has also implemented its own AI-powered tool that guides customers by comparing their body shape with garment dimensions. These technologies are already helping some brands reduce return rates significantly.
Beyond sizing, AI is also improving warehouse operations to prevent shipping mistakes. Smart cameras on order pickers’ trolleys at logistics firms help ensure the right product is selected, while AI-equipped robots track stock levels, reducing errors that lead to returns. As online shopping continues to grow, retailers hope these innovations will streamline processes and boost efficiency.
For more information on these topics, visit diplomacy.edu.
Baidu has announced plans to open-source its upcoming Ernie 4.5 chatbot series, citing inspiration from the success of AI firm DeepSeek. The move is aimed at encouraging greater adoption of its technology by making it more accessible.
Baidu also plans to provide premium chatbot services free of charge starting in April, with the open-source launch scheduled for June.
DeepSeek’s open-source approach, combined with its innovative R1 model, has gained widespread attention and increased adoption of foundational AI tools. Baidu CEO Robin Li highlighted the significance of this strategy in shaping the company’s decision.
He expressed confidence in the technology behind Ernie 4.5, which Baidu claims will be its most advanced model to date.
Despite advancements in AI, Baidu’s Q4 2024 results revealed a 2% revenue decline to 34.12 billion yuan, slightly ahead of expectations. Growth in the Cloud AI sector, up 26% year-on-year, partly offset a 7% drop in its online marketing business.
Weakness in China’s economy, particularly the property market, has contributed to reduced advertising demand.
Baidu handled 1.65 billion daily interactions on its Ernie platform in December, up from 600 million in August. However, its US-listed shares dropped 7% in early trading after the results.
For more information on these topics, visit diplomacy.edu.
A US federal judge has denied a request to temporarily block Elon Musk’s Department of Government Efficiency (DOGE) from accessing data from seven federal agencies or making further workforce cuts. The lawsuit, brought by 14 Democratic attorneys general, argued that DOGE was overstepping its authority by reshaping agencies and obtaining vast amounts of government information. However, Judge Tanya Chutkan ruled that the plaintiffs failed to prove immediate harm, allowing DOGE to continue operations.
Despite this decision, the judge acknowledged serious constitutional concerns regarding Musk’s authority. She noted that Musk had not been nominated by the US President Trump or confirmed by the Senate, raising potential violations of the Appointments Clause. In her ruling, Chutkan also criticised the Trump administration’s legal arguments, suggesting inconsistencies in its justification for DOGE’s powers.
While the restraining order was denied, the states can still pursue their case, potentially seeking a preliminary injunction to halt DOGE’s access to federal data. New Mexico Attorney General Raúl Torrez vowed to continue the legal fight, accusing Musk of destabilising government functions and acting without proper oversight. The battle over DOGE’s legitimacy is expected to intensify in the coming months.
For more information on these topics, visit diplomacy.edu.
Google Meet’s AI-driven note-taking feature is getting a major upgrade with the ability to generate action items from meeting transcripts. The update, powered by Google’s Gemini AI, will automatically identify key tasks, assign deadlines, and designate responsible individuals at the end of each meeting.
Originally launched in August 2024, the AI transcription tool already provides accurate speaker separation and structured summaries in Google Docs. With this latest enhancement, the technology aims to improve productivity by ensuring that key takeaways are actionable and well-organised.
The feature begins rolling out today but at a slower pace than usual, as Google closely monitors its performance and quality. While AI-generated notes can be a helpful time-saver, some users may still prefer manual control over meeting documentation, especially when handling sensitive information.
For more information on these topics, visit diplomacy.edu.
Former OpenAI chief technology officer Mira Murati has launched a new AI startup called Thinking Machines Lab, backed by a team of around 30 researchers and engineers from companies such as OpenAI, Meta, and Mistral. The startup aims to create AI systems that encode human values and address a wider range of applications than existing rivals, according to a blog post from the company.
Murati’s new venture demonstrates her ability to attract top talent, with two-thirds of the team made up of former OpenAI employees. Among them are Barret Zoph, a well-known researcher who joined Murati in leaving OpenAI in September, and John Schulman, OpenAI’s co-founder and the startup’s chief scientist. Schulman previously left OpenAI for Anthropic to focus on AI alignment, a key goal of Thinking Machines Lab.
The company’s approach differentiates itself by combining research and product teams in the design process. Thinking Machines Lab plans to contribute to AI alignment research by sharing code, datasets, and model specifications. Murati, now CEO of the startup, has previously played a major role in developing ChatGPT, and her exit from OpenAI reflects a broader trend of high-profile departures amid changes at the company.
For more information on these topics, visit diplomacy.edu.
Freelance marketplace Fiverr is rolling out new AI tools designed to help gig workers streamline their work and maintain competitiveness. The ‘Personal AI Creation Model’ allows freelancers to train AI systems on their own work, offering clients either automated or blended content.
Freelancers retain ownership of the generated content and can set prices for its use, with the service costing $25 per month.
The tools aim to address challenges in the gig economy, where generative AI has increased competition while reducing opportunities. Fiverr highlights that the AI creation model prioritises freelancers’ control, ensuring their data is not misused.
Additionally, a ‘Personal AI Assistant’ is available to help with routine tasks and client communication, priced at $29 per month or included with Fiverr’s premium subscription.
Fiverr is also introducing a programme granting company shares to its top-performing freelancers. While the specifics remain undisclosed, the initiative reflects efforts to support gig workers in an increasingly competitive market.
As generative AI reshapes industries, Fiverr’s initiatives could help freelancers navigate the evolving landscape while enhancing their earning potential and safeguarding their creative work.
For more information on these topics, visit diplomacy.edu.
Milan prosecutors have announced plans to drop a case against Google’s European division after the company agreed to settle a tax dispute by paying 326 million euros (£277 million). The settlement covers the period from 2015 to 2019, including penalties, sanctions, and interest.
The tax dispute stemmed from allegations that Google had failed to file and pay taxes on revenue generated in Italy, based on the digital infrastructure it operates within the country. This comes after the company settled a previous tax case with Italian authorities in 2017 by paying 306 million euros, which acknowledged Google’s permanent presence in Italy.
In 2023, Italy had requested that Google pay 1 billion euros in unpaid taxes and penalties. However, with this latest settlement, the case against the tech giant appears to be resolved for now.
For more information on these topics, visit diplomacy.edu.
South Korea has unveiled plans to acquire 10,000 high-performance GPUs by the end of the year, aiming to strengthen its national AI computing capabilities.
Officials emphasised that global AI competition has escalated from corporate rivalries to clashes between nations’ innovation ecosystems. Acting President Choi Sang-mok stated that public-private cooperation would be key to launching services at the country’s national AI computing centre.
The initiative comes amid the US government’s newly implemented restrictions on exporting AI chips, which categorise countries into tiers. South Korea remains exempt from these restrictions, allowing it access to advanced GPUs.
Details on the budget, specific GPU models, and private sector partners are expected to be finalised by September, according to the Ministry of Science and ICT.
US-based Nvidia dominates the GPU market with an 80% share, with rivals Intel and AMD trailing behind. Meanwhile, OpenAI, supported by Microsoft, is developing its first in-house AI chip to reduce dependency on Nvidia.
Chinese startup DeepSeek has also entered the AI scene with models prioritising computational efficiency over raw processing power, potentially narrowing the technological gap with US-made processors.
As nations race to secure technological supremacy, South Korea’s move signals its commitment to advancing its AI infrastructure and staying competitive in the global AI ecosystem.
For more information on these topics, visit diplomacy.edu.
Apple is set to introduce AI-driven features to its Vision Pro headset, including a new spatial content app and an enhanced guest user mode. The update, expected in April, may be available in beta for developers as early as this week, according to sources cited by Bloomberg.
The AI integration, dubbed ‘Apple Intelligence,’ will bring tools such as Writing Tools, Genmojis, and the Image Playground app to the Vision Pro. This move aligns with Apple’s broader push to incorporate AI into its ecosystem, as the company also explores AI partnerships for iPhones in China.
The Vision Pro, which launched in February 2024 with a hefty price tag of $3,499, initially saw strong sales but has since faced slowing demand. By enhancing its software with AI and new content features, Apple hopes to reignite consumer interest and better compete with more affordable alternatives, such as Meta’s Quest headsets.
For more information on these topics, visit diplomacy.edu.
Samsung Electronics has nominated two of its top chip executives and a semiconductor expert for its board, aiming to enhance its position in the competitive chip market. The move reflects a strategic effort to revitalise its struggling semiconductor business.
Jun Young-hyun, head of Samsung’s chip division, and Chief Technology Officer Song Jai-hyuk have been proposed for board membership. Additionally, Lee Hyuk-jae, a professor at Seoul National University and leader of its semiconductor research centre, has been named as an outside director.
Samsung has been working to regain its footing in the semiconductor industry, where it recently lost market leadership in high-bandwidth memory (HBM) chips to rival SK Hynix. HBM chips are crucial for AI-driven GPUs, a rapidly growing market dominated by firms like Nvidia.
The nominations will be presented for approval at Samsung’s shareholders’ meeting scheduled for 19 March. The company hopes the addition of these experts will strengthen its long-term strategy in the chip sector.
For more information on these topics, visit diplomacy.edu.
