Home / Blog / Hardware
Hardware วิเคราะห์จากสเปค + รีวิว

Analyze and review: ChatGPT Computer History tracks your clicks and keystrokes Analyze and review: ChatGPT Computer History tracks your clicks and keystrokes

Analyze the functionality, privacy, security, and implications of the ChatGPT Computer History feature that tracks clicks and keystrokes. Analyze the functionality, privacy, security, and implications of the ChatGPT Computer History feature that tracks clicks and keystrokes.

TL;DR: Computer History helps AI (artificial intelligence) build better context from clicks and typing, but users should carefully review the data scope, controls, and history deletion options. If it were me, I’d recommend enabling it only when working with non-sensitive information.

Computer History helps AI see more continuous usage context, so it may answer questions based on what users do on their computers more accurately than before. But this convenience comes with click data, typing data, and sensitive activity.

The key point is that users should understand what data is collected, what it is used for, and how it can be controlled or deleted. Greater accuracy does not mean the system will correctly understand intent every time, especially when switching between tasks or when personal information gets mixed in.

What users will see when it is enabled

When Computer History is enabled, the system displays a window requesting permission to track activity on the computer, including clicks, typing, and the windows or apps currently in use. Users should clearly review the scope before granting permission.

After it is enabled, the system will reference past activity when answering questions. Users may see relevant activity lists or context on the screen. Pay attention to app names, the content being recorded, and the time periods covered by the data.

What users will see when it is enabled

When Computer History is enabled, the system displays a window requesting permission to track activity on the computer, including clicks, typing, and the windows or apps currently in use. Users should clearly review the scope before granting permission.

After it is enabled, the system will reference past activity when answering questions. Users may see relevant activity lists or context on the screen. Pay attention to app names, the content being recorded, and the time periods covered by the data.

When the computer cannot remember what we just did

Imagine switching between multiple windows, searching several websites, and leaving a form partially completed. When you return to ask AI something, you have to explain every step again from the beginning, losing both time and concentration.

If ChatGPT can remember the sequence of clicks and typing, ongoing work may become smoother because you do not have to repeat the context. But the important question is whether this convenience is worth the risks associated with the stored data, especially when entering personal or important information. If Computer History were compared to a notebook, it would be a notebook that remembers the details of your workflow—but you would still need to be careful about who has permission to read it.

When the computer cannot remember what we just did

Imagine switching between multiple windows, searching several websites, and leaving a form partially completed. When you return to ask AI something, you have to explain every step again from the beginning, losing both time and concentration.

If ChatGPT can remember the sequence of clicks and typing, ongoing work may become smoother because you do not have to repeat the context. But the important question is whether this convenience is worth the risks associated with the stored data, especially when entering personal or important information.

Where Computer History fits into the ChatGPT ecosystem

Memory remembers information that users intentionally share, while screen sharing allows AI to see what is happening at that moment. Meanwhile, the mode that lets AI act on the user’s behalf will click or perform tasks according to instructions.

Computer History sits somewhere in the middle because it is a continuous data layer that stores context from clicks and typing to help ChatGPT understand the workflow that came before. This feature therefore moves ChatGPT away from being a chat that simply waits for questions and closer to an assistant that observes work continuously. But that convenience comes with questions about scope and privacy.

Where Computer History fits into the ChatGPT ecosystem

Memory remembers information that users intentionally share, while screen sharing allows AI to see what is happening at that moment. Meanwhile, the mode that lets AI act on the user’s behalf will click or perform tasks according to instructions.

Computer History sits somewhere in the middle because it is a continuous data layer that stores context from clicks and typing to help ChatGPT understand the workflow that came before. This feature therefore moves ChatGPT away from being a chat that simply waits for questions and closer to an assistant that observes work continuously. But that convenience comes with questions about scope and privacy.

From providing context yourself to having a usage history

Factor Traditional chatComputer History
What ChatGPT can perceive Only the text in the chatClicks and typing during work
How users must provide context Explain it themselves each timeThe system builds context from usage history
Continuity between tasks Often have to start explaining againCan connect ongoing tasks more effectively
Risks from stored data Limited to what is sentCovers more usage behavior
Level of control and history deletion Straightforward to controlRequires careful review of the scope and deletion options

The turning point is that users no longer have to explain everything themselves, but they must allow the system to see more of their workflow. The convenience therefore raises questions about what data should be stored and how much users can delete or stop the tracking.

From providing context yourself to having a usage history

Factor Traditional chatComputer History
What ChatGPT can perceive Only the text in the chatClicks and typing during work
How users must provide context Explain it themselves each timeThe system builds context from usage history
Continuity between tasks Often have to start explaining againCan connect ongoing tasks more effectively
Risks from stored data Limited to what is sentCovers more usage behavior
Level of control and history deletion Straightforward to controlRequires careful review of the scope and deletion options

Computer History is therefore like an older sibling that can manage ongoing work more effectively than traditional chat, the younger sibling. But the trade-off is greater exposure of usage data.

TL;DR: Computer History helps AI (artificial intelligence) build better context from clicks and typing, but users should carefully review the data scope, controls, and history deletion options. If it were me, I’d recommend enabling it only when working with non-sensitive information.

Computer History helps AI see more continuous usage context, so it may answer questions based on what users do on their computers more accurately than before. But this convenience comes with click data, typing data, and sensitive activity.

The key point is that users should understand what data is collected, what it is used for, and how it can be controlled or deleted. Greater accuracy does not mean the system will correctly understand intent every time, especially when switching between tasks or when personal information gets mixed in.

What users will see when it is enabled

When Computer History is enabled, the system displays a window requesting permission to track activity on the computer, including clicks, typing, and the windows or apps currently in use. Users should clearly review the scope before granting permission.

After it is enabled, the system will reference past activity when answering questions. Users may see relevant activity lists or context on the screen. Pay attention to app names, the content being recorded, and the time periods covered by the data.

What users will see when it is enabled

When Computer History is enabled, the system displays a window requesting permission to track activity on the computer, including clicks, typing, and the windows or apps currently in use. Users should clearly review the scope before granting permission.

After it is enabled, the system will reference past activity when answering questions. Users may see relevant activity lists or context on the screen. Pay attention to app names, the content being recorded, and the time periods covered by the data.

When the computer cannot remember what we just did

Imagine switching between multiple windows, searching several websites, and leaving a form partially completed. When you return to ask AI something, you have to explain every step again from the beginning, losing both time and concentration.

If ChatGPT can remember the sequence of clicks and typing, ongoing work may become smoother because you do not have to repeat the context. But the important question is whether this convenience is worth the risks associated with the stored data, especially when entering personal or important information. If Computer History were compared to a notebook, it would be a notebook that remembers the details of your workflow—but you would still need to be careful about who has permission to read it.

When the computer cannot remember what we just did

Imagine switching between multiple windows, searching several websites, and leaving a form partially completed. When you return to ask AI something, you have to explain every step again from the beginning, losing both time and concentration.

If ChatGPT can remember the sequence of clicks and typing, ongoing work may become smoother because you do not have to repeat the context. But the important question is whether this convenience is worth the risks associated with the stored data, especially when entering personal or important information.

Where Computer History fits into the ChatGPT ecosystem

Memory remembers information that users intentionally share, while screen sharing allows AI to see what is happening at that moment. Meanwhile, the mode that lets AI act on the user’s behalf will click or perform tasks according to instructions.

Computer History sits somewhere in the middle because it is a continuous data layer that stores context from clicks and typing to help ChatGPT understand the workflow that came before. This feature therefore moves ChatGPT away from being a chat that simply waits for questions and closer to an assistant that observes work continuously. But that convenience comes with questions about scope and privacy.

Where Computer History fits into the ChatGPT ecosystem

Memory remembers information that users intentionally share, while screen sharing allows AI to see what is happening at that moment. Meanwhile, the mode that lets AI act on the user’s behalf will click or perform tasks according to instructions.

Computer History sits somewhere in the middle because it is a continuous data layer that stores context from clicks and typing to help ChatGPT understand the workflow that came before. This feature therefore moves ChatGPT away from being a chat that simply waits for questions and closer to an assistant that observes work continuously. But that convenience comes with questions about scope and privacy.

From providing context yourself to having a usage history

Factor Traditional chatComputer History
What ChatGPT can perceive Only the text in the chatClicks and typing during work
How users must provide context Explain it themselves each timeThe system builds context from usage history
Continuity between tasks Often have to start explaining againCan connect ongoing tasks more effectively
Risks from stored data Limited to what is sentCovers more usage behavior
Level of control and history deletion Straightforward to controlRequires careful review of the scope and deletion options

The turning point is that users no longer have to explain everything themselves, but they must allow the system to see more of their workflow. The convenience therefore raises questions about what data should be stored and how much users can delete or stop the tracking.

From providing context yourself to having a usage history

Factor Traditional chatComputer History
What ChatGPT can perceive Only the text in the chatClicks and typing during work
How users must provide context Explain it themselves each timeThe system builds context from usage history
Continuity between tasks Often have to start explaining againCan connect ongoing tasks more effectively
Risks from stored data Limited to what is sentCovers more usage behavior
Level of control and history deletion Straightforward to controlRequires careful review of the scope and deletion options

Computer History is therefore like an older sibling that can manage ongoing work more effectively than traditional chat, the younger sibling. But the trade-off is greater exposure of usage data.

TL;DR: Computer History helps AI (artificial intelligence) build better context from clicks and typing, but users should carefully review the data scope, controls, and history deletion options. If it were me, I’d recommend enabling it only when working with non-sensitive information.

Computer History helps AI see more continuous usage context, so it may answer questions based on what users do on their computers more accurately than before. But this convenience comes with click data, typing data, and sensitive activity.

The key point is that users should understand what data is collected, what it is used for, and how it can be controlled or deleted. Greater accuracy does not mean the system will correctly understand intent every time, especially when switching between tasks or when personal information gets mixed in.

What users will see when it is enabled

When Computer History is enabled, the system displays a window requesting permission to track activity on the computer, including clicks, typing, and the windows or apps currently in use. Users should clearly review the scope before granting permission.

After it is enabled, the system will reference past activity when answering questions. Users may see relevant activity lists or context on the screen. Pay attention to app names, the content being recorded, and the time periods covered by the data.

What users will see when it is enabled

When Computer History is enabled, the system displays a window requesting permission to track activity on the computer, including clicks, typing, and the windows or apps currently in use. Users should clearly review the scope before granting permission.

After it is enabled, the system will reference past activity when answering questions. Users may see relevant activity lists or context on the screen. Pay attention to app names, the content being recorded, and the time periods covered by the data.

When the computer cannot remember what we just did

Imagine switching between multiple windows, searching several websites, and leaving a form partially completed. When you return to ask AI something, you have to explain every step again from the beginning, losing both time and concentration.

If ChatGPT can remember the sequence of clicks and typing, ongoing work may become smoother because you do not have to repeat the context. But the important question is whether this convenience is worth the risks associated with the stored data, especially when entering personal or important information. If Computer History were compared to a notebook, it would be a notebook that remembers the details of your workflow—but you would still need to be careful about who has permission to read it.

When the computer cannot remember what we just did

Imagine switching between multiple windows, searching several websites, and leaving a form partially completed. When you return to ask AI something, you have to explain every step again from the beginning, losing both time and concentration.

If ChatGPT can remember the sequence of clicks and typing, ongoing work may become smoother because you do not have to repeat the context. But the important question is whether this convenience is worth the risks associated with the stored data, especially when entering personal or important information.

Where Computer History fits into the ChatGPT ecosystem

Memory remembers information that users intentionally share, while screen sharing allows AI to see what is happening at that moment. Meanwhile, the mode that lets AI act on the user’s behalf will click or perform tasks according to instructions.

Computer History sits somewhere in the middle because it is a continuous data layer that stores context from clicks and typing to help ChatGPT understand the workflow that came before. This feature therefore moves ChatGPT away from being a chat that simply waits for questions and closer to an assistant that observes work continuously. But that convenience comes with questions about scope and privacy.

Where Computer History fits into the ChatGPT ecosystem

Memory remembers information that users intentionally share, while screen sharing allows AI to see what is happening at that moment. Meanwhile, the mode that lets AI act on the user’s behalf will click or perform tasks according to instructions.

Computer History sits somewhere in the middle because it is a continuous data layer that stores context from clicks and typing to help ChatGPT understand the workflow that came before. This feature therefore moves ChatGPT away from being a chat that simply waits for questions and closer to an assistant that observes work continuously. But that convenience comes with questions about scope and privacy.

From providing context yourself to having a usage history

Factor Traditional chatComputer History
What ChatGPT can perceive Only the text in the chatClicks and typing during work
How users must provide context Explain it themselves each timeThe system builds context from usage history
Continuity between tasks Often have to start explaining againCan connect ongoing tasks more effectively
Risks from stored data Limited to what is sentCovers more usage behavior
Level of control and history deletion Straightforward to controlRequires careful review of the scope and deletion options

The turning point is that users no longer have to explain everything themselves, but they must allow the system to see more of their workflow. The convenience therefore raises questions about what data should be stored and how much users can delete or stop the tracking.

From providing context yourself to having a usage history

Factor Traditional chatComputer History
What ChatGPT can perceive Only the text in the chatClicks and typing during work
How users must provide context Explain it themselves each timeThe system builds context from usage history
Continuity between tasks Often have to start explaining againCan connect ongoing tasks more effectively
Risks from stored data Limited to what is sentCovers more usage behavior
Level of control and history deletion Straightforward to controlRequires careful review of the scope and deletion options

Computer History is therefore like an older sibling that can manage ongoing work more effectively than traditional chat, the younger sibling. But the trade-off is greater exposure of usage data.

TL;DR: Computer History helps AI (artificial intelligence) build better context from clicks and typing, but users should carefully review the data scope, controls, and history deletion options. If it were me, I’d recommend enabling it only when working with non-sensitive information.

Computer History helps AI see more continuous usage context, so it may answer questions based on what users do on their computers more accurately than before. But this convenience comes with click data, typing data, and sensitive activity.

The key point is that users should understand what data is collected, what it is used for, and how it can be controlled or deleted. Greater accuracy does not mean the system will correctly understand intent every time, especially when switching between tasks or when personal information gets mixed in.

What users will see when it is enabled

When Computer History is enabled, the system displays a window requesting permission to track activity on the computer, including clicks, typing, and the windows or apps currently in use. Users should clearly review the scope before granting permission.

After it is enabled, the system will reference past activity when answering questions. Users may see relevant activity lists or context on the screen. Pay attention to app names, the content being recorded, and the time periods covered by the data.

What users will see when it is enabled

When Computer History is enabled, the system displays a window requesting permission to track activity on the computer, including clicks, typing, and the windows or apps currently in use. Users should clearly review the scope before granting permission.

After it is enabled, the system will reference past activity when answering questions. Users may see relevant activity lists or context on the screen. Pay attention to app names, the content being recorded, and the time periods covered by the data.

When the computer cannot remember what we just did

Imagine switching between multiple windows, searching several websites, and leaving a form partially completed. When you return to ask AI something, you have to explain every step again from the beginning, losing both time and concentration.

If ChatGPT can remember the sequence of clicks and typing, ongoing work may become smoother because you do not have to repeat the context. But the important question is whether this convenience is worth the risks associated with the stored data, especially when entering personal or important information. If Computer History were compared to a notebook, it would be a notebook that remembers the details of your workflow—but you would still need to be careful about who has permission to read it.

When the computer cannot remember what we just did

Imagine switching between multiple windows, searching several websites, and leaving a form partially completed. When you return to ask AI something, you have to explain every step again from the beginning, losing both time and concentration.

If ChatGPT can remember the sequence of clicks and typing, ongoing work may become smoother because you do not have to repeat the context. But the important question is whether this convenience is worth the risks associated with the stored data, especially when entering personal or important information.

Where Computer History fits into the ChatGPT ecosystem

Memory remembers information that users intentionally share, while screen sharing allows AI to see what is happening at that moment. Meanwhile, the mode that lets AI act on the user’s behalf will click or perform tasks according to instructions.

Computer History sits somewhere in the middle because it is a continuous data layer that stores context from clicks and typing to help ChatGPT understand the workflow that came before. This feature therefore moves ChatGPT away from being a chat that simply waits for questions and closer to an assistant that observes work continuously. But that convenience comes with questions about scope and privacy.

Where Computer History fits into the ChatGPT ecosystem

Memory remembers information that users intentionally share, while screen sharing allows AI to see what is happening at that moment. Meanwhile, the mode that lets AI act on the user’s behalf will click or perform tasks according to instructions.

Computer History sits somewhere in the middle because it is a continuous data layer that stores context from clicks and typing to help ChatGPT understand the workflow that came before. This feature therefore moves ChatGPT away from being a chat that simply waits for questions and closer to an assistant that observes work continuously. But that convenience comes with questions about scope and privacy.

From providing context yourself to having a usage history

Factor Traditional chatComputer History
What ChatGPT can perceive Only the text in the chatClicks and typing during work
How users must provide context Explain it themselves each timeThe system builds context from usage history
Continuity between tasks Often have to start explaining againCan connect ongoing tasks more effectively
Risks from stored data Limited to what is sentCovers more usage behavior
Level of control and history deletion Straightforward to controlRequires careful review of the scope and deletion options

The turning point is that users no longer have to explain everything themselves, but they must allow the system to see more of their workflow. The convenience therefore raises questions about what data should be stored and how much users can delete or stop the tracking.

From providing context yourself to having a usage history

Factor Traditional chatComputer History
What ChatGPT can perceive Only the text in the chatClicks and typing during work
How users must provide context Explain it themselves each timeThe system builds context from usage history
Continuity between tasks Often have to start explaining againCan connect ongoing tasks more effectively
Risks from stored data Limited to what is sentCovers more usage behavior
Level of control and history deletion Straightforward to controlRequires careful review of the scope and deletion options

Computer History is therefore like an older sibling that can manage ongoing work more effectively than traditional chat, the younger sibling. But the trade-off is greater exposure of usage data.