In partnership with

Welcome to AI Entrepreneurs

In this issue, we dive into the latest AI advancements reshaping industries—from customer support powered by Decagon and OpenAI’s GPT models, to Nvidia’s cutting-edge robotics network, HOVER, transforming control in humanoid robots. We also explore groundbreaking strides in AI hardware with OpenAI’s proprietary chip collaboration, OpenAI's new voice mode for desktop, and Stable Diffusion’s more accessible model for creative AI.

Whether you’re curious about healthcare innovations, ethical marketing practices, or the future of conversational AI, this edition brings you insights into how AI is setting new standards across fields.

Stay tuned for exciting insights into the future of AI!

🤖 AI in Customer Support

Decagon and OpenAI Power Fully Automated Customer Service

Decagon, launched in 2023, has rapidly become a leader in fully automated customer support for companies like Curology, BILT, and Duolingo. With OpenAI’s advanced models, Decagon delivers seamless and fast customer interactions, handling millions of inquiries for businesses without the need for human agents. Using a combination of GPT models, such as GPT-3.5 and GPT-4, Decagon’s AI agents can manage entire customer lifecycles, offering speed and accuracy across global support networks.

Decagon's platform automates up to 91% of customer support for major clients, ensuring high-quality service. By refining OpenAI models and conducting evaluations, Decagon remains at the forefront of AI advancements, enhancing customer satisfaction with faster, context-aware responses.

Decagon integrates Anthropic’s Claude to provide scalable, personalized 24/7 support, automating interactions for businesses like Eventbrite and Rippling. This platform reduces costs, enhances satisfaction, and allows human agents to focus on complex tasks, ensuring efficient, human-like AI support.

Implications:
Decagon's AI-driven support solutions are setting a new standard in customer service, advancing speed, accuracy, and scalability across industries. This shift not only enhances customer experiences but also enables companies to deliver cost-effective, quality service at scale.

AI in Hardware 💻

OpenAI Teams Up with Broadcom & TSMC for Proprietary AI Chip Development

OpenAI is taking a bold step into AI hardware by collaborating with Broadcom and TSMC to develop its first proprietary chip, aiming to enhance the performance of its AI models while managing infrastructure demand. This partnership marks a shift toward custom-built AI solutions tailored to OpenAI's needs, which currently rely on Nvidia and AMD chips. Broadcom will aid in the chip’s development, while TSMC provides manufacturing capabilities, setting OpenAI up for increased efficiency and reduced reliance on external chip suppliers.

With rapidly growing infrastructure requirements, OpenAI had explored options like building its own chip factories but ultimately opted against it due to high costs and time constraints. Instead, the company will concentrate on in-house chip design, ensuring a stable and scalable supply chain. This move aligns with OpenAI’s strategy to future-proof its AI systems with enhanced, cost-effective hardware solutions designed to scale its cutting-edge models.

AI in Robotics 🤖

Nvidia’s HOVER: Compact Neural Network Transforms Robot Control

Nvidia researchers have unveiled HOVER, a lightweight neural network with only 1.5 million parameters, designed to control humanoid robots using multiple input methods—from VR headsets to motion capture. Unlike traditional control systems that are resource-intensive, HOVER achieves advanced movement control efficiently, outperforming specialized systems tailored for singular input types.

Trained in Nvidia’s Isaac simulation environment, HOVER enables a year’s worth of robotic training in just 50 minutes on one GPU. This system transitions smoothly from virtual simulation to real-world application, handling robot movements without additional fine-tuning. It accepts versatile input sources, including XR headsets, motion capture, RGB cameras, and joysticks.

Implications :
With its compact yet powerful design, HOVER offers scalable, efficient robotic control that can easily adapt across various robot models. This advance could reduce the complexity and cost of humanoid robot systems, accelerating development in fields like telemedicine, industrial automation, and remote assistance.

Introducing Roku Ads Manager: CTV Ads Made Simple

We have entered a new era of TV advertising where the barriers to entry are lower than ever, and Roku Ads Manager is leading the charge. Growth marketers now have access to Roku’s audience reach in the US. with a self-service ad platform that’s built for performance. Roku powers 47% of all TV streaming time* in the US, so your brand can run ads alongside premium content and meet your audience where they’re already engaged. *Comscore CTV Report, 2024

Reach your audience on Roku

💡Spotlight

Are Your AI Marketing Practices Ethical? (Honestly.)

Join top AI experts on November 12th to learn how leading companies are implementing AI ethically and effectively. Featuring WPP's Chief AI Officer, UK Government's AI Policy Lead, and real case studies from marketing leaders who've done it successfully.

Save Your SpotI

AI in Conversation 🗣️

ChatGPT Goes Vocal on Desktop: Advanced Voice Mode Arrives

OpenAI has launched the Advanced Voice Mode for ChatGPT on macOS and Windows desktop apps, bringing natural, real-time conversational abilities to the desktop environment. Initially available on mobile, this feature now lets users interact with ChatGPT as if speaking to a person, enabling smooth interruptions and even recognizing emotional cues, making the interaction more dynamic and human-like.

The addition of voice mode enhances the ChatGPT experience, transforming traditional text exchanges into lifelike interactions across platforms. With this upgrade, OpenAI aims to create a seamless and engaging experience for desktop users, taking voice-powered AI to new levels of usability and personalization.

The Hustle is a must-read for innovators.

Their daily newsletter delivers the latest stories in business and tech -- what to learn from them, and how to capitalize on them. For daily insights straight to your inbox, join 2.5M+ innovators who start their day with The Hustle.It’s 100% free and can be consumed in just 5 minutes.

Sign up today.

AI in Art 🎨

Stability AI Unveils Stable Diffusion 3.5 Medium for Accessible Image Generation

Stability AI has officially released Stable Diffusion 3.5 Medium, a model designed with lower VRAM requirements, making advanced image generation more accessible for users with standard consumer hardware. Available on HuggingFace, this model boasts an adapted architecture compared to the Stable Diffusion 3.5 Large and Large Turbo versions, making it a more efficient alternative for users seeking high-quality AI imagery without the high computational cost.

With 2.5 billion parameters, Stable Diffusion 3.5 Medium can generate images from 0.25 to 2 megapixels, making it versatile for a range of use cases. Stability AI offers this model for free across research, non-commercial, and commercial purposes for users or companies with annual revenue below $1 million. This move is set to broaden AI-powered creativity, supporting creators with scalable, cost-effective tools.

AI in Healthcare 🏥

AIHealthTech Insider: Issue #20

Dive into AIHealthTech Insider Issue #20 to uncover the latest breakthroughs.

This issue explores AI's impact on patient care and diagnostics, featuring Eye-AD's dementia detection via retinal imaging and NewYork-Presbyterian's $2 billion AI initiative. Highlights include Aidoc and NVIDIA's BRIDGE framework for seamless AI adoption in healthcare, and Color Health's AI copilot, powered by OpenAI, which speeds up cancer treatment.

Startup Success: Instant Access to AI Entrepreneur Courses!