¶ Introduction and overview
What if artificial intelligence agents could not only browse the web but also handle your business tasks seamlessly? Welcome to The AI Agent Daily Brief, your go-to for the latest AI updates. Today is Monday, February 24, 2025. Here’s what you need to know about how AI agents are starting to automate the enterprise. Let’s dive in.
¶ OpenAI's Operator and Orby's LAMs and their business potential
So far, 2025 has been all about AI agents, specifically how generative AI technology is being used to automate actions. We’ve seen OpenAI’s Operator debut, showcasing a prototype agent that can browse the web and perform tasks for you. Now, there’s a new player in town called Orby, and they’re bringing this innovative approach to the enterprise world with a type of AI model they call a Large Action Model, or LAM.
I recently had the chance to speak with Orby’s co-founder and Chief Technology Officer, Will Lu. Before co-founding Orby, Lu was an engineering leader at Google Cloud AI, so he’s no stranger to cutting-edge technology. Large Action Models, or LAMs, are designed to automate complex tasks within an enterprise setting. Imagine an AI agent that doesn’t just manage your calendar but can also handle customer inquiries, process orders, and even make decisions based on real-time data.
The potential to transform how businesses operate is enormous. Lu explained that the key to LAMs is their ability to learn from a wide variety of data sources and execute tasks that traditionally require human judgment. This isn’t just about efficiency; it’s about enabling businesses to scale operations without the proportional increase in human resources. The implications of LAMs for the future of work are significant.
As companies like Orby continue to refine these AI models, we could see a shift where human workers are managing teams of AI agents, each specialized in different aspects of business operations. It’s a fascinating development that could reshape industries across the board.
¶ Google and Salesforce partnership and future integrations
Google and Salesforce are joining forces to bring more options to customers in the world of artificial intelligence agents. They've just announced an expansion of their partnership, which means that joint customers will now have access to Google Gemini within Salesforce’s Agentforce platform. This collaboration aims to provide more customer choice by allowing AI agents to leverage Gemini’s capabilities to process images, audio, and video.
Imagine an AI that can recognize error codes in photos or detect emotions in voice recordings. That's what they're aiming for, and it's pretty exciting stuff. The companies are confident that this partnership will provide customers with crucial flexibility. It allows them to develop customized AI solutions without being locked into a single model provider.
Srini Tallapragada, Salesforce’s President and Chief Engineering and Customer Success Officer, highlighted how this deepened partnership gives customers choice in the applications and models they want to use. He emphasized that Salesforce offers a comprehensive agentic AI platform, while Google Cloud provides some of the most powerful and capable AI models and development tools available. Together, they’re creating a robust environment for businesses to scale with digital labor.
Let’s unpack some of the additional features this partnership will deliver. First up, the power of choice. Multimodal AI capabilities mean that agents can process images, audio, and text, enabling them to perform advanced tasks. Enhanced context and reasoning come from Gemini’s two-million-token context window, which allows agents to reference vast data sets like entire codebases or years of customer interactions.
And, of course, there’s greater speed and efficiency, thanks to Google’s TPUs and advanced processing techniques, delivering real-time responses that boost operational efficiency and cut costs. Thomas Kurian, CEO of Google Cloud, explained that this integration allows enterprise customers to deploy some of their most critical applications on Google’s secure, AI-optimized infrastructure with minimal friction.
Customers have been asking for a seamless experience across Salesforce and Google Cloud, and this partnership aims to meet that demand by accelerating AI transformations with top-tier AI models, data analytics, and more. Later this year, we can expect an even stronger integration between Salesforce Service Cloud and Google Cloud Customer Engagement Suite. This will create a seamless and intelligent support experience.
Imagine real-time voice translation and sentiment analysis, where language barriers are broken down, and tone and audio cues are analyzed for better sentiment understanding. Plus, intelligent agent handoffs will allow virtual agents to transition complex interactions smoothly to Agentforce in Service Cloud, ensuring multi-step resolutions are handled with ease.
The partnership also strengthens the integration between Salesforce’s Data Cloud and Google BigQuery, enabling agents to access real-time information and improve decision-making. For instance, in supply chain management, an Agentforce-powered AI could track shipments, monitor inventory, and anticipate disruptions using real-time data on weather, port congestion, and geopolitical events.
This kind of integration ensures enhanced security and trust with features like dynamic grounding, zero data retention, and toxicity detection via the Einstein Trust Layer. Once Salesforce products launch on Google Cloud, businesses will have the option to purchase them through the Google Cloud Marketplace, making it even easier to access these powerful tools.
It’s clear that Google and Salesforce are not just expanding their partnership; they’re reshaping how businesses can utilize AI to transform their operations. It’s a fascinating time to watch these tech giants collaborate and push the boundaries of what’s possible with AI agents.
¶ AI advancements in the financial sector with Westpac and Accenture
Westpac is making waves in the financial sector by teaming up with Accenture to bring artificial intelligence agents into the fold. These aren’t just your run-of-the-mill chatbots; they’re designed to mimic human decision-making and perform actions autonomously, a concept known as agentic AI. It’s an exciting leap forward from traditional generative AI, which primarily surfaces and communicates data without taking action.
AI agents that not only assist with software programming but could eventually help bankers process loan applications. It’s a vision of the future where AI doesn’t just support human workers but actively participates in the decision-making process, enhancing efficiency and accuracy. Accenture’s collaboration with Westpac marks a significant step in digital transformation within the banking industry.
By leveraging agentic AI, Westpac aims to streamline operations, reduce processing times, and ultimately provide better service to their customers. This isn’t just about cutting costs; it’s about reimagining the role of technology in banking. The potential applications of these AI agents extend beyond programming and loans. Imagine them analyzing complex data sets to provide insights for strategic decision-making or even predicting market trends to advise on investment strategies.
The possibilities are as vast as they are exciting. This partnership underscores a broader trend in the financial services industry: the shift towards digital transformation and innovation. As banks like Westpac continue to explore the capabilities of AI, we’re likely to see a wave of new tools that enhance customer experience and operational efficiency across the board. What’s clear is that Westpac and Accenture are setting the stage for a future where AI agents play a crucial role in banking.
As these technologies evolve, they’ll not only transform how banks operate but also how they interact with customers, making financial services more accessible and personalized.
¶ Microsoft's AI agents in Dynamics 365 and market competition
Microsoft's AI agents are here to take over your routine business tasks, but the question is, can they be trusted to deliver? In a push to automate business operations, Microsoft rolled out these autonomous AI agents as part of their Dynamics 365 platform. These agents are engineered to handle everything from sales to supply chain management, aiming to automate business processes without human intervention or coding expertise.
Think about it like this: Microsoft’s AI agents aren’t just glorified chatbots. They’re designed to function with minimal supervision, executing complex workflows across various sectors. Microsoft even compared these agents to microservices, each performing specialized tasks, much like how an organization uses its existing models to build effective agents. But can these agents truly stand out in a crowded AI landscape? That's the million-dollar question.
Microsoft hopes that by surpassing traditional AI offerings, their AI agents will have the edge over competitors like Salesforce, IBM, and Nvidia. They’re banking on the agents’ ability to reason over data and context to make judgments. Some companies, like McKinsey & Co., have already tested Microsoft’s AI agents, reporting impressive results. They managed to reduce lead times by ninety percent and cut administrative tasks by thirty percent with the help of these agents.
It’s a glimpse into how these AI tools might revolutionize business operations. Yet, not everyone is sold. Gartner research shows that many businesses are still in the early testing phases of Microsoft’s Copilot AI, with adoption rates lagging. Concerns around hallucinations, input accuracy, and ethical issues like bias and trust continue to loom large. Meanwhile, the competition is fierce.
Salesforce, for example, launched Agentforce, a platform that allows businesses to customize AI agents for various functions. Their CEO, Marc Benioff, has been vocal, criticizing Microsoft’s Copilot as lacking sophistication, even calling it "Clippy 2.0." IBM is also in the race with their Granite 3.0 8B models, focusing on advanced reasoning and structured workflows. With such robust competition, Microsoft has to prove that their AI solutions offer tangible value.
So, the real test for Microsoft’s AI agents will be their performance in real-world business scenarios. Will they fulfill the tech industry’s bold claims, or will they end up as another Silicon Valley experiment that fails to catch on? The coming months will be crucial in determining whether these autonomous agents are truly revolutionary or just another tech gimmick. As the market evolves, businesses are becoming more strategic about AI adoption.
It’s no longer a question of whether to adopt AI but how to do it effectively. The focus is on solutions that integrate seamlessly with existing workflows while being mindful of cost and complexity. Microsoft has a window of opportunity to invest and capture the market, but it’s a race against time and competition. With strategic implementation and addressing current concerns, these AI agents could indeed change the landscape of business automation.
¶ Alexandr Wang's insights on AI pods and job market impacts
Imagine a future where instead of doing the tasks themselves, office workers become managers of their own "pods" of artificial intelligence agents. Alexandr Wang, the co-founder of Scale AI, envisions this scenario as we dive deeper into the age of agentic AI. He recently shared his insights in a podcast interview with Theo Von, suggesting that people won’t be out of work but rather promoted to overseeing groups of AI agents.
Wang, a 28-year-old billionaire, described it as almost everyone getting a promotion to a managerial role. Instead of directly handling tasks, employees would manage a "pod" of 10 AI agents, ensuring they perform tasks correctly and without errors. This shift, he believes, will redefine the traditional office job landscape. The fear of labor disruption due to AI is real, with studies like the one from the Brookings Institute highlighting potential risks to office and administrative jobs.
High-paying fields such as STEM, business, and finance have been flagged as vulnerable to AI displacement. The study noted that more than 30 percent of workers could see at least half of their tasks disrupted by AI technology. However, Wang is optimistic that AI’s integration into these industries will be balanced by new job opportunities. As AI grows, he foresees a surge in roles centered around improving, controlling, and being part of AI systems.
This isn’t just about losing jobs but transforming the nature of work itself. Scale AI, which Wang describes as the "Uber for AI," is poised to be at the forefront of creating these new job opportunities. The company specializes in providing the crucial data tech companies need to refine their AI systems, employing gig workers globally to label information accurately. Despite facing some legal challenges, Scale AI continues to attract major clients like the U.S.
Department of Defense, OpenAI, Google, and Meta. Wang emphasized that outside of Scale AI’s network, the broader labor market will see new roles focused on regulating AI. Managing AI systems will become crucial to ensure they don’t deviate from intended functions. This shift is part of a larger historical trend where technological advancements reshape the job market.
He drew parallels to past technological shifts, such as the agricultural revolution, where technology made farming more efficient, enabling people to explore new opportunities. Wang believes the rise of AI will similarly open doors to unexplored job roles and redefine how we perceive work.
¶ Conclusion and sign-off
That’s it for today’s The AI Agent Daily Brief. From Google and Salesforce's partnership to Westpac's innovative use of AI, and now Scale AI's vision of a future workforce managing pods of AI agents, it’s clear that AI is reshaping industries in profound ways. Thanks for tuning in—subscribe to stay updated. This is Michelle, signing off. Until next time.