Articles for category: AI News

March 4, 2025

ikayaniaamirshahzad@gmail.com

Humanities teaching will have to adapt to AI | Artificial intelligence (AI)

I agree with Prof Andrew Moran and Dr Ben Wilkinson (Letters, 2 March) that cheap and easy‑to‑use AI tools create problems for universities, but the reactions of many academics to these new developments remind me of the way some people responded to the arrival of cheap pocket calculators in the 1970s. Reports of the imminent death of maths teaching in schools proved exaggerated. Maths teachers had to adapt, not least to teach students the longstanding rule “garbage in, garbage out”; if students had no idea of the fundamental principles and ideas behind maths, they would not realise their answer was meaningless.

March 4, 2025

ikayaniaamirshahzad@gmail.com

Salesforce Launches AgentExchange Marketplace for AI Agents

Salesforce on Tuesday launched AgentExchange, a marketplace for Agentforce, allowing partners, developers, and the Agentblazer community to build and monetise AI components. The platform supports businesses in the $6 trillion digital labour market. AgentExchange includes over 200 partners, such as Google Cloud, Docusign, and Box, providing prebuilt solutions for AI agents. Businesses can discover, test, and purchase prebuilt actions, topics, and templates on the marketplace or within Salesforce’s agent-building tools. AgentExchange is now live at AgentExchange.Salesforce.com. Prompt templates and topics can be listed and packaged immediately, while agent templates will be available for listing in April 2025. AgentExchange builds on

March 4, 2025

ikayaniaamirshahzad@gmail.com

Is AI progress slowing down?

By Arvind Narayanan, Benedikt Ströbl, and Sayash Kapoor. After the release of GPT-4 in March 2023, the dominant narrative in the tech world was that continued scaling of models would lead to artificial general intelligence and then superintelligence. Those extreme predictions gradually receded, but up until a month ago, the prevailing belief in the AI industry was that model scaling would continue for the foreseeable future. Then came three back-to-back news reports from The Information, Reuters, and Bloomberg revealing that three leading AI developers — OpenAI, Anthropic, and Google Gemini — had all run into problems with their next-gen models.

March 4, 2025

ikayaniaamirshahzad@gmail.com

WilsonAI raises $1.7M in pre-seed funding for ‘AI paralegal’

WilsonAI Co., the provider of an artificial intelligence-powered assistant for in-house legal operations, said today it has raised $1.7 million in pre-seed funding to accelerate the development of what the company calls an AI paralegal that companies can “hire” for their teams. The company was founded in 2024 by Chief Executive Gus Neate, formerly a lawyer at British multinational law firm Clifford Chance LLP, and Chief Technology Officer Alex Wang, formerly a developer at New York City-based investment firm D. E. Shaw & Co. It wants to change how legal teams operate using AI. “We’re building an AI that doesn’t

March 4, 2025

ikayaniaamirshahzad@gmail.com

Fearing AI will take their jobs, California workers plan a long battle against tech

Meanwhile, people like Amba Kak see opportunities for gains by workers against technological threats but said that it may require strategically picking the right battles. Kak previously advised the Federal Trade Commission and is executive director of the AI Now Institute, a nonprofit that researches the human rights impact of the technology.   — Kak told CalMatters she plans to pay more attention to activity in state legislatures in places like California and New York, where lawmakers are already considering a bill that protects people from AI in a manner similar to California’s to Senate Bill 1047, a controversial bill requiring AI safeguards that

March 4, 2025

ikayaniaamirshahzad@gmail.com

Less is more: How ‘chain of draft’ could cut AI costs by 90% while improving performance

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More A team of researchers at Zoom Communications has developed a breakthrough technique that could dramatically reduce the cost and computational resources needed for AI systems to tackle complex reasoning problems, potentially transforming how enterprises deploy AI at scale. The method, called chain of draft (CoD), enables large language models (LLMs) to solve problems with minimal words — using as little as 7.6% of the text required by current methods while maintaining or even improving accuracy. The findings were published in a

March 4, 2025

ikayaniaamirshahzad@gmail.com

Opera introduces browser-integrated AI agent

Opera has introduced “Browser Operator,” a native AI agent designed to perform tasks for users directly within the browser. Rather than acting as a separate tool, Browser Operator is an extension of the browser itself—designed to empower users by automating repetitive tasks like purchasing products, completing online forms, and gathering web content. Unlike server-based AI integrations which require sensitive data to be sent to third-party servers, Browser Operator processes tasks locally within the Opera browser. Opera’s demonstration video showcases how Browser Operator can streamline an everyday task like buying socks. Instead of manually scrolling through product pages or filling out

March 4, 2025

ikayaniaamirshahzad@gmail.com

What’s the best AI model to handle $1 million in freelance software engineering?

OpenAI recently introduced SWE-Lancer, a benchmark that tests how well today’s most advanced AI models can handle real-world software engineering tasks. Even though they just tweeted about it 19 hours ago, this paper is already the top paper on AIModels.fyi in the agents category and growing fast. Their paper presents a benchmark that evaluates AI language models on 1,488 actual freelance software engineering tasks from Upwork, collectively worth $1 million in payouts. These tasks were sourced from the open-source Expensify repository. Fig 2. “Evaluation flow for SWE manager tasks; during proposal selection, the model has the ability to browse the