openai, deepseek and chatgpt Updates

OpenAI is making significant moves in the financial sector, reportedly hiring over 100 former investment bankers for a project codenamed 'Project Mercury.' These professionals, sourced from major firms like JPMorgan Chase, Goldman Sachs, and Morgan Stanley, are being paid $150 per hour to train AI models, including enhancements for ChatGPT, on complex financial modeling tasks. The initiative aims to automate routine entry-level banking jobs, such as building financial models for IPOs and buyouts, and refining presentations. This development raises questions about the future of employment for junior bankers. Meanwhile, in the broader AI landscape, PolyAI showcased advanced AI voice agents, and discussions at Bloomberg Tech in London focused on the evolving human-machine connection with insights from ICONIQ and Jaimee. SoundPatrol is tackling AI copyright issues with neural fingerprinting, while Lakeview City Council is set to debate AI policy. Roofing Matrix has launched an AI sales tool for contractors, and Cequence Security is bolstering its channel program for AI security growth. Globally, leaders in Hong Kong discussed AI's impact on education and business, stressing the need for agility and human literacy alongside technological fluency. In a cryptocurrency trading competition, DeepSeek AI showed promising returns, outperforming GPT-5, though the event highlighted AI's limitations in understanding market volatility and the need for human oversight. Starbucks' CEO also hinted at AI integration for voice ordering and new menu items, alongside existing AI tools like Green Dot Assist.

Key Takeaways

• OpenAI has hired over 100 former investment bankers for 'Project Mercury' to train AI on financial modeling, aiming to automate entry-level banking tasks.
• These ex-bankers are paid $150 per hour to help train OpenAI's models, including ChatGPT, for financial tasks.
• PolyAI demonstrated advanced AI voice agents at Bloomberg Tech in London.
• Discussions at Bloomberg Tech also explored the evolving relationship between humans and AI, with insights from ICONIQ and Jaimee.
• SoundPatrol is developing neural fingerprinting to combat AI copyright infringement.
• Roofing Matrix launched an AI sales tool to help roofing contractors automate lead follow-up and booking.
• Cequence Security is enhancing its channel program to drive growth in application, API, and AI security.
• Starbucks' CEO hinted at AI-powered voice ordering and new menu items, with the company already using AI tools like Green Dot Assist.
• In a crypto trading competition, DeepSeek AI achieved a 20% return, while GPT-5 lost 36%, highlighting AI's challenges with market volatility.
• Global leaders in Hong Kong emphasized the importance of balancing technological fluency with human literacy in the age of AI.

OpenAI hires 100+ ex-bankers for AI financial modeling

OpenAI has hired over 100 former investment bankers from major firms like JPMorgan Chase, Morgan Stanley, and Goldman Sachs for a secret initiative called Project Mercury. This project aims to train AI systems on financial modeling, enabling them to analyze data, generate reports, and potentially offer investment advice. The recruitment highlights OpenAI's focus on using specialized human expertise to create more advanced AI for the financial industry. While OpenAI has not confirmed details, the hiring suggests a significant investment in this area.

OpenAI pays ex-bankers $150/hour for AI financial training

OpenAI is paying over 100 former investment bankers $150 per hour to train its AI on financial modeling tasks through a project called 'Project Mercury.' These professionals, including ex-employees from JPMorgan Chase, Goldman Sachs, and Morgan Stanley, are feeding financial models into AI to expand its practical use in sectors like finance. The application process involves an AI chatbot interview and financial modeling tests. This initiative aims to automate repetitive tasks in investment banking, though it raises concerns about job security for junior bankers.

OpenAI's Project Mercury aims to automate banking jobs

OpenAI is reportedly developing 'Project Mercury,' an initiative to automate entry-level banking tasks using artificial intelligence. The project involves hiring former investment bankers to train AI models for tasks like financial modeling and data analysis. This move aims to enable AI to perform jobs traditionally done by junior bankers. The development raises questions about the future employment landscape in the financial industry as AI capabilities grow.

OpenAI hires 100 ex-bankers for secret Wall Street automation project

OpenAI has hired over 100 former investment bankers from firms like JPMorgan Chase, Morgan Stanley, and Goldman Sachs for a secret project called 'Project Mercury.' These individuals are paid $150 an hour to train AI models to perform time-consuming investment banking tasks, such as building financial models for IPOs and buyouts. The project aims to automate the routine work handled by entry-level analysts. Contractors also test AI outputs and receive early access to OpenAI's financial modeling tools.

OpenAI uses ex-bankers to train ChatGPT on financial models

OpenAI is reportedly developing 'Project Mercury' to enhance ChatGPT's capabilities for the banking industry. Over 100 former investment bankers are being paid $150 per hour to help train OpenAI's models. Their work involves creating financial models and refining PowerPoint slides, tasks typically done by junior bankers. This initiative aims to make ChatGPT more commercially valuable for businesses, particularly in finance.

PolyAI showcases advanced AI voice agents

PolyAI demonstrated its next-generation AI voice agents at Bloomberg Tech in London. Shawn Wen, Co-Founder and Chief Technology Officer at PolyAI, provided a live demo of the technology. The showcase highlighted the future capabilities of AI in voice interactions.

AI leaders discuss human-machine connection at Bloomberg Tech

AI leaders shared insights on the evolving relationship between humans and machines at Bloomberg Tech in London. Lauren Kunze of ICONIQ and Sreyna Rath of Jaimee discussed how emotionally intelligent AI is reshaping these connections. Their conversation with Bloomberg's Parmy Olson focused on the future of human-machine interaction.

SoundPatrol fights AI copyright issues with neural fingerprinting

SoundPatrol is developing a solution to combat AI copyright infringement using neural fingerprinting. Co-founders Walter De Brouwer and Hollywood agent Michael Ovitz discussed their startup's technology with CNBC's Julia Boorstin. This method aims to make media identifiable even when distorted by AI, protecting creators' rights.

Lakeview Council to debate AI policy

The Lakeview City Council will discuss a new policy regarding the use of artificial intelligence at its upcoming meeting. The council will consider Resolution 25-03, which aims to establish guidelines for AI implementation. Other agenda items include filling a council vacancy and updates on local projects.

Roofing Matrix launches AI sales tool for contractors

Roofing Matrix has launched its AI Sales Team, a new platform designed to help roofing contractors close more deals. The AI tool automates lead follow-up, qualification, and appointment booking 24/7. Early users have reported significant increases in lead-to-appointment conversion rates and new revenue. Roofing Matrix offers the AI Sales Team on a done-for-you basis with a performance guarantee.

Cequence Security boosts channel program for AI security growth

Cequence Security is enhancing its channel program to drive growth and profitability in application, API, and AI security. The company is transitioning to a channel-only model with structured tiers, improved enablement, and profitability incentives. This program aims to help partners address API security complexity and limited expertise by providing unified protection and managed services. Cequence Security has already onboarded several strategic partners.

Starbucks CEO hints at AI voice ordering and new menu items

Starbucks CEO Brian Niccol revealed potential changes, including AI-powered voice ordering and new menu items, possibly including alcoholic beverages. The company is experimenting with AI for forecasting, scheduling, and voice-to-vision ordering, aiming for a seamless customer experience. Starbucks has also implemented Green Dot Assist, an AI tool to help baristas, and Smart Queue to manage orders.

Global leaders meet in Hong Kong on AI's impact on education and business

Global leaders convened in Hong Kong to discuss thriving in the age of AI, focusing on education and business transformation. Northeastern University President Joseph E. Aoun emphasized balancing technological fluency with human literacy. Speakers highlighted the need for agility and innovation in adapting to AI's rapid changes. The event explored how AI, intergenerational collaboration, and human literacy are reshaping industries and learning.

AI trading competition reveals machine intellect's limits

The Nof1.ai tournament showcased six advanced AI models, including DeepSeek, Grok, and GPT-5, trading cryptocurrencies with initial stakes of $10,000. DeepSeek achieved a 20% return, while Grok earned 13%, but GPT-5 lost 36%. The competition highlighted AI's difficulty in understanding market emotions and volatility, revealing the need for human oversight in trading. The event also raised questions about regulatory responsibility for AI trading losses.

Market volatility, China trade, and AI job impact analyzed

The past week saw significant market volatility with fluctuating equity indexes despite numerous risky headlines. Discussions focused on US-China trade relations, with potential de-escalation talks ahead of a meeting between President Trump and Xi Jinping. The article also touches on Micron Technologies' potential exit from the Chinese market and Nebius Group's new AI cloud product, 'Aether.' It raises questions about which jobs are most vulnerable to AI automation.

Sources

• OpenAI hires over 100 ex-investment bankers to train AI on financial modeling
• OpenAI Is Paying Ex-Investment Bankers $150 an Hour to Train Its AI
• AI to take over banking jobs? OpenAI launches ‘Project Mercury’ for automation of entry-level tasks
• Why OpenAI is hiring 100 ex-bankers: Inside the ChatGPT-maker's secret project to automate Wall Street's grunt work
• OpenAI has an army of ex-investment bankers making financial models to train ChatGPT
• PolyAI Demos the Future of AI Voice Agents
• AI Leaders on the Future of Human and Machine Connection
• Neural fingerprinting makes media identifiable even when distorted by AI: SoundPatrol co-founders
• Lakeview City Council to discuss policy regarding artificial intelligence
• Roofing Matrix Launches AI Sales Team to Help Contractors Close More Jobs
• Cequence Security Enhances Channel Program to Accelerate Growth and Drive Profitability Across the Application, API, and AI Security Ecosystem
• AI voice ordering, new menu items teased by Starbucks CEO as possible changes
• Global leaders meet in Hong Kong to share playbook for thriving in the age of AI
• The AI Trading Duel Unveiling the Fragility of Machine Intellect
• Surveying a Wild Week, China Trade Latest, Who Loses Jobs First to AI