The Future Of AI Agents With Dharmesh Shah

HubSpot's Memorable Moments and AI Updates

HubSpot celebrated its 18th anniversary in June.
CEO shares moments like the launch of his son 13 years ago and the launch of HubSpot CRM 10 years ago.
HubSpot has launched generative AI products, including ChatSpot and Breeze Copilot.
The CEO mentions the growing popularity of AI agents and jokes about the trend.
AI models are getting bigger and more capable, with larger neural networks and improved reasoning and writing abilities.
AI models are also growing exponentially faster than Moore's Law predicted.
There are multiple highly functional AI models, each excelling at different tasks.
OpenAI is expected to release a new impressive model called Strawberry.

OpenAI launched a new model called GPT o1, which is known for its impressive reasoning abilities and the ability to share its thought process.
The rapid advancement of AI models allows for the creation of more sophisticated AI agents with diverse goals.
AI models now support multimodal inputs, including images, audio, and video, giving them a sense of sight.
AI agents utilize AI models and tools to accomplish goals that require multiple steps.
An AI agent can run autonomously and have executive planning capabilities, but these are not necessary to be considered an AI agent.
All AI agents have some common characteristics: they use AI models, can utilize tools, and have a memory to remember tasks.
Ongoing development in AI includes coding and late-night work to enhance capabilities and functionality.

Agent.AI is a professional network specifically designed for AI agents.
The purpose of this network is to create hybrid teams consisting of both humans and AI agents.
Agent.AI offers a platform where AI agents can have profiles, be followed, and be hired by users.
The network currently has agents in various fields such as marketing, sales, customer service, and operations.
The focus is on helping marketers generate traffic and qualified leads through conversion rate optimization.
The Conversion Rate Optimizer agent simplifies the process by providing expert suggestions for website optimization.
The agent analyzes webpages and provides recommendations to increase conversions.
The agent is easy to use and can be applied to any public webpage.
By utilizing AI agents, sales reps can save time on research and spend more time with customers.

Company Research Agent:
Saves time by consolidating information from multiple sources.
Provides a multi-page report with sections on founders, recent news, web traffic, organic keywords, and more.
Allows users to ask custom questions and provides answers by crawling the company's website and the internet.
Custom questions can be saved and answered in future reports.
Integration with HubSpot Smart CRM allows for easy addition of companies and access to CRM functionalities.
Offers tracking capabilities to monitor important updates and notify users.
Executive Briefing Agent:
Reduces the work of creating an executive briefing for customer success managers.
Pulls information from the HubSpot Smart CRM, including past calls and issues.
Utilizes the Company Research Agent to gather additional data from the web.
Generates a concise executive briefing that can be converted into an audio file for easy listening.
Provides an efficient way for customer success managers to stay updated on their accounts. Note: The bullet points have been adjusted to summarize the main benefits of the two agents mentioned in the provided text.

Prediction that most people will be using agents this year.
Many people will be building agents this year.
Anyone can build an AI agent.
Agent Builder allows users to easily build agents through five simple steps.
Access to various data sources and models.
Agent Builder has already gained 47,000 users, making it a successful innovation project.
1,700 people have signed up for the waitlist for Agent Builder.
Users do not need coding skills, just curiosity and a few easy clicks.
The process involves building a user interface and utilizing an API to connect with HubSpot.

The Agent User Interface (AUI) allows agents to talk to each other and collaborate.
Agent Builder automatically includes an AUI for every agent built on the platform.
HubSpot believes that the next generation customer platform should have a smart CRM with unified structured and unstructured data.
The core CRM should have foundational AI built-in.
HubSpot aims to make it exceptionally easy to use, build, share, and distribute agents.
The vision for the future is "there's an agent for that" for every marketing, sales, and customer service use case.
HubSpot expects there to be thousands of agents available, mostly built by partners and the community.
Agents and humans will be connected on a network and can collaborate in a vibrant marketplace.
Agents are seen as future digital teammates, not just tools.
AI does not reduce the value of humans, but rather raises it.
AI may be faster and smarter, but it can't be more human because it lacks human experiences.