{"id":12473,"date":"2025-04-28T07:41:07","date_gmt":"2025-04-28T14:41:07","guid":{"rendered":"https:\/\/www.salesforce.com\/?p=12473"},"modified":"2025-08-06T14:54:24","modified_gmt":"2025-08-06T14:54:24","slug":"future-of-ai-agents","status":"publish","type":"post","link":"https:\/\/www.salesforce.com\/eu\/blog\/future-of-ai-agents\/","title":{"rendered":"Think AI Agents Are Impressive Now? Just Wait"},"content":{"rendered":"\n

You\u2019ve seen the ripples of what AI agents<\/a> can do. Now, get ready for the tidal wave.\u00a0<\/p>\n\n\n\n

The future of AI agents? It\u2019s agents that can hear your commands and see the world alongside you. It’s AI orchestrating AI, creating a symphony of digital labour<\/a>. It’s agents with digital memories that learn your patterns and foresee your next move. Forget incremental upgrades.\u00a0The AI agent of tomorrow will be a force multiplier unleashed.<\/p>\n\n\n\n

Here are five ways that agents \u2014 like those powered by Agentforce<\/a>, the Salesforce platform for building AI agents \u2014 will evolve.<\/p>\n\n\n\n

1. Agents will understand much more than text <\/h2>\n\n\n\n

Today, artificial intelligence (AI)<\/a>\u00a0agents\u00a0are primarily text-based: They interact with and process information in the form of typed messages, commands or queries. But they can\u2019t understand or respond to images, audio or video and this lack of contextual richness limits their ability to handle complex, real-world situations.\u00a0\u00a0<\/p>\n\n\n\n

\u201cWe can do a lot more than that. The world is full of information that’s not words,\u201d said Juan Carlos Niebles, research director at Salesforce. \u201cAn agent will be able to look at an image or listen to an audio clip or a person\u2019s voice. You give it eyes and ears and you open a door to other types of data that the agent could understand and process.\u201d<\/p>\n\n\n\n

In field service<\/a>, a service technician could use their Agentforce app to record audio of a sputtering car engine and ask Agentforce to diagnose the problem. In marketing, AI agents could analyse social media trends in real time, like a sudden spike in videos about \u201980s-inspired leg warmers. In customer service, they could analyse audio from service calls, picking up on recurring negative sentiment from the tone of a customer\u2019s voice.\u00a0<\/p>\n\n\n\n

To do this, AI agents must be able to see images and videos and hear sounds. In a nutshell, this is accomplished by adding neural networks that can translate the input data from the new modality (voice, audio, video, visual) into tokens that a large language model <\/a>(LLM) can understand. <\/p>\n\n\n\n

To explore the future of multimodal AI, Salesforce AI Research has developed xGen-MM-Vid, a multimodal LLM that helps AI interpret videos. Take a look. <\/p>\n\n\n\n

\n

<\/p>\n<\/div><\/div>\n\n\n\n

<\/p>\n\n\n\n

The future of AI agents is also voice: Soon, they’ll understand and respond to our spoken requests. Imagine verbally asking an agent to do the following: <\/p>\n\n\n\n

\u201cAccess Q2 sales performance from the shared drive and analyse KPIs with a focus on revenue growth, churn rate and customer acquisition cost. Then, summarise the analysis, present your top findings and identify the most important areas for improvement. Finally, recommend two actionable next steps to remedy the challenges.\u201d <\/p>\n\n\n\n

This is an enormous time-saver, but that\u2019s only half the story. Since AI agents can quickly process huge amounts of data, they can help workers uncover trends they might otherwise miss, identifying subtle but important correlations between data. <\/p>\n\n\n\n

2. Agents will work effortlessly with other agents <\/h2>\n\n\n\n

Right now, AI agents generally work solo to handle individual tasks, like customer service or inventory management. But soon, systems of multiple agents, each with a specialised role, will work together to carry out more complex tasks to achieve broader goals. This will be a force multiplier in terms of scale, speed and strategic decision-making, fundamentally altering operational workflows and enabling a level of co-ordinated efficiency that\u2019s not possible with an all-human workforce. Call it the age of agent-to-agent or A2A. <\/p>\n\n\n\n

For example, in ecommerce, a multi-AI agent system could include one agent handling customer enquiries, another managing inventory and a third optimising pricing based on-demand, all working together in real time to improve efficiency and maximise sales.<\/p>\n\n\n\n

This collaboration could be taken even further, with AI working across organisations, industries and entire ecosystems to negotiate, optimise, inform and maybe even influence decisions on a global scale. <\/p>\n\n\n\n

\"\"\n\t\t\t\n\t\t\t\t\n\t\t\t<\/svg>\n\t\t<\/button>
A single AI agent tackles one job at a time, while multiple specialised agents work as a team, each handling part of a larger, more complex project.<\/em>
<\/figcaption><\/figure>\n\n\n\n

Imagine a global supply chain where AI agents from different companies work together to co-ordinate logistics, source materials and manage distribution. They react in real time to delays, demand shifts and disruptions, keeping everything running smoothly. The result? A self-regulating, ultra-efficient network that practically runs itself.<\/p>\n\n\n\n

But even in an increasingly autonomous A2A world, human oversight will remain critical for setting ethical boundaries, addressing unforeseen circumstances and ensuring alignment with business strategy and human values. The role of humans may evolve to focus on strategic guidance and ensuring the AI ecosystem serves human needs and goals.<\/p>\n\n\n\n

These interdependent, multi-agent systems will require sophisticated co-ordination, orchestration and governance \u2014 by both humans and AI \u2014 that transcends traditional oversight models. Experts in Salesforce AI Research and the Global AI Practice recently laid out three methods of managing<\/a> multi-agent complexity.<\/p>\n\n\n\n

3. Orchestrator agents will manage teams of agents <\/h2>\n\n\n\n

Just like human workers, teams of AI agents need a manager to direct and co-ordinate different activities and oversee multistep tasks through to competition. But in the future of AI agents, those managers will be other AI agents. <\/p>\n\n\n\n

Silvio Savarese, executive president and chief scientist, AI Research, at Salesforce, recently described<\/a> an orchestrator agent that co-ordinates the work of multiple specialist agents, similar to how a restaurant\u2019s general manager oversees the work of hosts, servers, managers, chefs, cooks and expediters.  <\/p>\n\n\n\n

What does this look like in an enterprise context? A service agent could process a customer\u2019s enquiry while an inventory agent checks product availability. A logistics agent calculates delivering and a billing agent reviews and processes payment options. <\/p>\n\n\n\n

\"\"\n\t\t\t\n\t\t\t\t\n\t\t\t<\/svg>\n\t\t<\/button>
An orchestrator AI agent manages a team of specialised agents.<\/figcaption><\/figure>\n\n\n\n

As Savarese noted, \u201cThe orchestrator agent co-ordinates all these inputs into a coherent, effective, on-brand and contextually relevant response for the human at the helm to review, refine and share with the customer.\u201d <\/p>\n\n\n\n

In other words, instead of juggling multiple individual agents, employees can work with one smart lead agent that co-ordinates everything behind the scenes. <\/p>\n\n\n\n

And orchestrator agents make it easy to grow and adapt. Think of them as smart connectors for your AI. When your business changes or you find new AI tools, you can easily add them to your system without having to rebuild everything from scratch. This keeps your AI set-up ready for the future and able to handle new things.<\/p>\n\n\n\n

4. Agents will get much better at reasoning <\/h2>\n\n\n\n

A reasoning engine<\/a> is an AI system that mimics human-like decision-making and problem-solving based on certain rules, data and logic. Essentially, it\u2019s how an agent decides what actions to take and which data is needed to take those actions.\u00a0<\/p>\n\n\n\n

A strong reasoning engine allows an AI to go beyond surface-level data, analysing complex patterns, inferring user intent and understanding the nuances of a user’s behaviour, preferences and needs. Instead of just reacting to keywords or simple commands, it can grasp the underlying context of a user’s interactions.  <\/strong><\/p>\n\n\n\n

\u201cThis is about getting the AI agent to do more complex things. It\u2019s not just a one-step task,\u201d said Niebles. \u201cIt\u2019s a chain of maybe 10 different actions to accomplish a goal.\u201d <\/p>\n\n\n\n

For example, here\u2019s how an AI reasoning engine could power a multipronged, multistep marketing campaign:<\/p>\n\n\n\n