Ace, the latest computer control agent developed by General Agents Co., has set a new benchmark in terms of speed and accuracy. Compared to existing agents like Operator from OpenAI, Ace demonstrates remarkable performance, executing tasks significantly faster and with higher precision.

One of the key advantages of Ace is its lightning-fast execution speed. Whether it's copying an image from Google Chrome and sharing it on iMessage, zipping and uploading files to Google Drive, or even controlling video editing software like Premiere Pro, Ace can complete these tasks in a fraction of the time it would take a human. The agent's ability to navigate various applications and perform complex operations with such speed is truly impressive.

Moreover, Ace's accuracy is also noteworthy, with a click accuracy of nearly 80% in the "Ace Control Medium" benchmark. This level of precision is crucial for ensuring reliable and consistent task completion, making Ace a valuable tool for a wide range of applications.

The agent's underlying model, which is trained on behavioral data rather than just text and images, is a key factor in its superior performance. This approach allows Ace to generalize better and leverage the wealth of data available from user interactions and task recordings, enabling it to learn and execute tasks more effectively.

As Ace continues to scale up its training resources, it is expected to become even more intelligent and capable, further solidifying its position as the fastest and most accurate computer control agent on the market.

Ace, the latest computer use agent, has demonstrated remarkable speed and accuracy in a variety of tasks. One of the key advantages of Ace is its ability to outperform human users in common computer-based activities.

In the demonstration, Ace showcased its prowess in browsing the web, managing files, and even controlling video editing software. The agent was able to quickly and efficiently copy an image from Google Chrome, share it via iMessage, and zip and upload files to Google Drive - all at a pace far exceeding what a human user could achieve.

Furthermore, Ace demonstrated its proficiency in video editing by seamlessly slowing down a clip in Adobe Premiere Pro, splitting the clip, and exporting it as an MP4 file. The agent's ability to execute these complex tasks with such speed and accuracy is truly impressive.

The benchmarks provided further highlight Ace's superiority over other computer use agents, such as Operator from OpenAI. Ace's click accuracy and action prediction latency significantly outperform its competitors, making it a highly capable and efficient tool for a wide range of computer-based tasks.

As the agent's training resources are scaled up, Ace is expected to become even more intelligent and capable, further solidifying its position as a game-changer in the field of computer use agents.

Ace, the latest computer use agent, has demonstrated remarkable speed and accuracy in performing a wide range of tasks. One of the key advantages of Ace is its ability to efficiently research flights and manage contacts.

In the flight research example, Ace was able to quickly search for the cheapest flight from San Francisco to New York on May 4th, navigating through Google Flights and completing the task faster than a human could. This highlights Ace's efficiency in handling complex, multi-step processes.

Similarly, Ace showcased its prowess in contact management by seamlessly importing a list of contacts into HubSpot. The agent was able to add each contact one by one, accurately filling in the necessary fields, all at a remarkably fast pace.

These examples illustrate Ace's potential to streamline common workflows and save users valuable time. With its combination of speed and precision, Ace can significantly enhance productivity and efficiency in various domains, from travel planning to customer relationship management.

Ace, the latest computer use agent from General Agents Co., has demonstrated remarkable performance in various benchmarks, outshining its competitors. The key highlights are:

• Click Accuracy: Ace Control Medium achieves nearly 80% click accuracy, outperforming other models like Operator (OpenAI), Malmo 72b, CLA 3.7, Sonet UI, and more.
• Action Prediction Latency: Ace Control Small and Ace Control Medium have a fraction-of-a-second latency, compared to 6.3 seconds for Operator.
• Scalability: As Ace's training resources are scaled up, its performance continues to improve, showcasing the potential for even greater capabilities in the future.

These benchmarks highlight Ace's superior speed and accuracy, making it a standout performer in the field of computer use agents. The unique behavioral training paradigm employed by Ace, which leverages screen recordings, mouse, and keyboard logs, allows it to learn and generalize tasks more effectively than traditional language and vision models.

Ace, the latest computer use agent, employs a novel training approach that sets it apart from traditional language and vision models. Unlike models trained on text and images, Ace is trained on behavioral data - the process that generates text, images, and other work outputs.

This behavioral training paradigm offers several key advantages:

1. Generalization: Training on behavior, rather than just the end products, allows Ace to generalize better and apply its skills to a wider range of tasks.

2. Step-by-step Reasoning: The use of step-by-step reasoning in Ace's training further enhances its ability to break down and execute complex tasks.

3. Accessibility for Domain Experts: Domain experts can simply record themselves performing tasks using familiar tools, without the need to learn new processes. This unlocks a vast, untapped source of training data.

4. Efficiency: Ace is trained using a fraction of the computing resources required by larger language models, yet it demonstrates impressive speed and accuracy.

As Ace's training resources are scaled up, the model will continue to grow more intelligent and capable, unlocking new possibilities for efficient and versatile computer use. This unique approach to behavioral training sets Ace apart as a promising step forward in the field of intelligent agents.

Unlike traditional language and vision models trained on text and images, Ace is trained on behavioral data - the process that generates text, images, and other work outputs. This approach to training intelligent agents offers several key advantages:

1. Generalization: Training on behavioral data allows Ace to generalize better, as corroborated by the use of step-by-step reasoning in the training process.

2. Accessibility for Domain Experts: Creating behavioral data is more natural for domain experts, who can simply record themselves performing tasks using the tools they are already familiar with, without the need to learn new tools or processes.

3. Untapped Data Source: The wealth of screen recordings, mouse and keyboard logs, and other behavioral data that companies like Apple and Microsoft possess represents an untapped resource for training intelligent agents. By allowing users to opt-in and share this data, these companies could empower the development of more capable and efficient agents.

4. Efficient Training: Ace is able to achieve impressive performance using a fraction of the computing resources required by larger language models, demonstrating the potential of behavioral data-driven training approaches.

As Ace continues to scale up its training resources, it is poised to become increasingly intelligent and capable, showcasing the promising future of intelligent agents that embrace behavioral data as the foundation for their development.

Ace, the new computer use agent launched by General Agents Co., is a game-changer in the world of AI-powered automation. With its lightning-fast speed, unparalleled accuracy, and ability to perform a wide range of tasks, Ace outperforms existing solutions like Operator from OpenAI.

The key differentiator of Ace is its unique training approach, which focuses on learning from behavioral data rather than just text and images. This allows Ace to generalize better and adapt more effectively to various tasks and scenarios.

As the training resources are scaled up, Ace is poised to become even more intelligent and capable, showcasing the potential of this new behavioral training paradigm. The benchmarks presented demonstrate Ace's superior performance in terms of click accuracy and action prediction latency, making it a highly efficient and reliable tool for a variety of applications.

The examples showcased, from web browsing and file management to video editing and flight booking, highlight Ace's versatility and the time-saving benefits it can provide. With its seamless integration across platforms and the ability to control the user's computer, Ace offers a compelling solution for streamlining and automating various workflows.

Overall, Ace's launch represents a significant step forward in the field of AI-powered automation, and its continued development and scaling hold the promise of revolutionizing the way we interact with and leverage technology in our daily lives.