In 2024, AI really ushered in a period of rapid development in technology development and application, and AI giants have predicted that the singularity will arrive.

In 2025, AI agents will rise and reshape the job market! A number of AI company executives said that by 2025 AI agents will be able to do a variety of jobs.

Julian Rose went so far as to assert:

We are about to witness the greatest transfer of wealth in history.

By 2034, new AI agents will:

- Displacing 800 million jobs (McKinsey)

- Add $19.9 trillion to the global economy (IDC)

Your 9-to-5 job will disappear!

The impact of AI agents on employment

With the rapid development of AI, agents are transforming from assistive tools into "virtual employees" capable of completing complex tasks autonomously.

This shift not only improves work efficiency, but also has a huge impact on the traditional employment structure. Many jobs with high repeatability and clear rules are gradually being replaced by AI agents.

At the same time, the demand for innovative, strategic positions has increased, and this trend of AI agents replacing human labor has brought efficiency dividends, but also triggered profound thinking and discussion about occupational safety.

AI agents enter the workplace

Charles Lamanna, Corporate Vice President, Business & Industry Copilot Division, Microsoft:

By the end of 2025, you will have teams of agents working for you. These agents might include: IT agents that fix technical glitches before you know it; Supply chain agents that prevent disruptions while you sleep; Selling agents that break down barriers between business systems to track leads; And financial agents to close accounts more quickly and efficiently.

Megh Gautam, Chief Product Officer, Crunchbase:

By 2025, AI investment will shift from the experimental phase to the execution phase. Companies will move away from general-purpose AI applications in favor of solutions to specific, high-value business problems. We will see two main trends: First, in handling routine but complex operational tasks, AI agents will come to the fore. Second, the widespread use of AI tools has led to significant improvements in optimizing sales and customer support automation.

Scott Beechuk, Partner, Norwest Venture Partners:

By the end of 2025, about 20 percent of enterprise software buyers will interact with suppliers through AI. In 2024, a number of AI Sales Development Representative (SDR) products were launched, and most of these purchases will be operational by 2025. Their success will pave the way for the next generation of sales automation. The new generation of AI account managers will begin to appear at the end of 2025 and be fully promoted in 2026.

China Widener, Vice Chairman, Technology, Media and Telecommunications, Deloitte USA:

We're talking about agent AI - intelligent assistants that can handle tasks autonomously, such as solving customer problems, writing code, and detecting cyberattacks. Like those chatbots you communicate with, they're finally ready to graduate and join the workforce. By 2025, 25% of enterprises will begin using these intelligent assistants, marking a fundamental shift in who we work with and how we work with them.

Paul Drews, Managing Partner, Salesforce Ventures:

We are in the midst of a technological transformation: the transition from generative AI to agent AI. 2024 is the year of building and testing AI models, but agents are the next step in applying AI to the real world. Consumers should expect every business company they are associated with to create an agent. We will see them in areas such as banking, insurance, healthcare and retail. By the end of 2025, agents will be part of our digital lives.

Prediction of AI agents

Stefan Mesken, Vice President of Research at DeepL:

AI will not only understand users better, but will also proactively provide advice, collaborate meaningfully, and adapt to individual needs. These advanced personalization features have long existed, but are currently limited to researchers or developers. Working with AI will increasingly be like working with smart colleagues.

Brandon Roberts, Senior Vice president of People Analytics and AI at ServiceNow:

AI has the potential to increase productivity for most companies by 10 to 20 percent over the next three to five years. As the real benefits of AI are increasingly valued, more companies will invest in using AI to optimize the roles and skills they need.

Shawn Carolan, Partner, Menlo Ventures:

We will see native AI applications emerge in almost every large consumer category: voice interaction will replace traditional menu navigation, and AI-powered systems delivering instant personalized customer service responses will become the norm. AI will be more than just voice, it will have faces, facial expressions and more empathy, and human interaction with AI will enter a new phase.

Timothy Young, CEO of Jasper:

As AI gets deeper and deeper into systems and data, our relationship with AI will change: instead of giving AI orders, AI will give us suggestions, insights, and solutions that will change the way we make decisions at work and in life. Leaders must not only respond to technological change, but also foster trust and collaboration between people and AI.

Andy Sack, Forum3 co-founder:

Over the next year, consumer behavior will change radically. Ai-driven platforms such as Perplexity, ChatGPT and Google's SGE (Search Generation Experience) will redefine the way search is conducted. Instead of relying on a bunch of links, users will get direct answers and suggestions for action.

Yash Sheth, COO and co-founder of Galileo:

Multimodal AI will become a reality, and voice will become an increasingly common user interface. AI will become more integrated into our daily lives. More AI will be added to our everyday apps (Instagram, Spotify, Doordash) and the tools we use at work (Google Workspace, Salesforce, email).

Advantages of AI agents

AI agents are not just automated tools, they are autonomous systems capable of learning from every interaction, making complex decisions, and constantly improving themselves.

They are also far better at reading, writing, cooperating, and making decisions than humans:

The reason why IDC AI has a $19.9 trillion impact on the global economy is not an unfounded exaggeration, but because AI agents:

• Can work 24 hours a day a week without rest;

• Never make the same mistake;

• Making decisions 1,000 times faster than humans.

The effect of multiple AI agents working together is more amazing! Altman is very excited about this, saying: "AI intelligence can run a company!"

By 2030, McKinsey predicts that AI agents will replace 70% of office jobs.

The remaining 30 percent of office work may require uniquely human skills like strategic thinking, relationship-building, emotional intelligence, and more.

But overall, technology-driven and routine jobs are at greater risk of displacement than those in the creative industries and the public sector; The less adaptability a job requires, the more likely it is to be replaced.

Google is ready

Many netizens rediscovered the white paper on AI agents released by Google in April.

This white paper covers the basics of large language model agents and provides a brief introduction to the Langchain implementation.

The white paper address: https://www.kaggle.com/whitepaper-agents

Five key takeaways:

1. Agents are "enhanced" AI: By using tools, making plans, and taking actions, they go beyond basic models and become more like how humans solve problems.

2. Everything depends on the setup: the agent needs three parts to work: an intelligent model for decision-making, tools to interact with the real world, and a system to manage plans and actions.

3. Tools change everything: apis, databases, and real-time data give agents superpowers that allow them to not only answer questions, but actually perform tasks.

4. Building Agents becomes easier: Platforms like LangChain and Vertex AI simplify the creation and deployment of agents, making it easy to implement even complex tasks.

5. Still work to do: There are still some major challenges that need to be addressed before agents become truly widespread, such as improving ease of use, interoperability, and privacy friendliness.

The architecture and components of the agent

The white paper proposes to use frameworks such as LangChain and LangGraph to build agent architectures. By introducing reasoning frameworks such as ReAct, thought chains, and thought trees, AI agents can solve problems and make decisions more effectively.

In the white paper, Google's Vertex AI Agent Builder is also recommended, arguing that agents will go beyond generative AI models to become independent applications with the ability to combine reasoning, logic, and external information.

AI Agents are able to access real-time information and make practical recommendations through extensions, functions, and data storage at the tool layer. AI agents can manage complex business processes and perform tasks that are difficult to accomplish with traditional language models.

The white paper proposes the concept of "agent chain collaboration". Multiple specialized agents can work together, similar to the cooperation of human experts, with each agent contributing specific expertise to solve complex challenges together.

Musk warned years ago

Assume that one in seven employees loses their jobs, a large number of jobs are replaced by AI agents, and

• Entire industries disappear overnight;

• Careers built over decades become obsolete;

• The job market is in disarray.

If that happens, Musk's warning will become a prophecy.