Digital privacy and the inner workings of artificial intelligence are often shrouded in mystery, leading to common questions about how much these systems actually observe. When someone types the phrase "what i was doing" into an AI interface, it usually stems from one of two motivations: a curiosity about the AI's tracking capabilities or a moment of personal forgetfulness where the user hopes the technology can provide a digital breadcrumb trail. The reality is that modern AI assistants operate within very strict boundaries, and understanding these limits is essential for any user concerned about their data security.

Understanding the privacy boundaries of modern AI

Large Language Models (LLMs) and conversational AI assistants are designed to process information provided within a specific interaction. They do not function like a global operating system that records every keystroke or browser tab opened on your device. When you ask an AI what you were doing, the response is typically a clarification of its lack of access. This transparency is a cornerstone of responsible AI development, ensuring that users realize the machine is not a silent observer of their offline or external digital lives.

The myth of constant AI surveillance

The idea that an AI is "watching" your screen or "listening" to your environment is a common misconception fueled by science fiction and the complex nature of targeted advertising. While some software uses cookies or tracking pixels to monitor web behavior for marketing purposes, conversational AIs are generally "stateless" by default. This means that each new session or chat is a blank slate. Unless a user has specifically enabled a feature that allows the AI to "see" their screen—a feature that requires explicit permission and is usually tied to specific productivity tools—the AI is blind to anything happening outside the immediate chat window.

Privacy protocols prevent AI companies from tapping into your system's background processes. This is why, if you were browsing for shoes on one website and then opened a chat to ask, "What was I doing?" the AI would have no record of your shopping activity. It only knows what you have explicitly typed or uploaded within that specific environment.

How context windows and session limits work

To understand why an AI cannot recall your past actions, it is necessary to look at the concept of the "context window." The context window is the amount of information the AI can "hold in its mind" at any given time during a conversation. This includes the current prompts and the previous messages in that specific thread.

  1. Session Isolation: Most AI platforms treat each conversation thread as an isolated container. Information does not leak from a chat you had yesterday into the chat you are having today.
  2. Data Deletion: Many professional-grade AI tools offer options to turn off chat history, meaning the data is processed in real-time and not stored for future retrieval by the model.
  3. No OS Integration: Unlike a computer's "Activity Monitor" or "Task Manager," the AI is not integrated into the kernel of your operating system. It does not have the permissions required to query your CPU history or RAM usage to see what applications were running five minutes ago.

What was I doing from a linguistic perspective

Beyond the technical questions of privacy, "what I was doing" is a fundamental phrase in the English language. It utilizes the past continuous tense, a grammatical structure used to describe actions that were ongoing at a specific point in the past. This tense is essential for storytelling, providing context, and explaining interruptions.

Grammatical structure of the past continuous tense

The past continuous tense is formed by using the past tense of the verb "to be" (was/were) and the present participle of the main verb (the "-ing" form).

  • Structure: Subject + was/were + verb-ing.
  • Example: "I was doing my homework when the phone rang."

In this sentence, "was doing" indicates an action that started before the phone rang and likely continued after, or was interrupted by, the ringing. It sets the scene. Without this tense, English speakers would struggle to convey the overlapping nature of events in time.

Why we say was and not were with the pronoun I

A frequent point of confusion for English learners, and even some native speakers, is the conjugation of the verb "to be" in the past tense. While "were" is used for plural subjects (we, they) and the second person (you), "was" is the correct singular form for the first person (I) and the third person (he, she, it).

  • Correct: "I was doing..."
  • Incorrect: "I were doing..." (Unless using the subjunctive mood for hypothetical situations, such as "If I were doing my job correctly...")

Using "was" is the standard for factual, indicative statements about the past. In the phrase "what I was doing," you are asking about a specific, real event that occurred, requiring the standard singular conjugation.

Common scenarios for using the phrase what was I doing

The phrase is rarely used in isolation; it usually serves a specific communicative purpose. Whether it is a rhetorical question or a request for clarification, the context changes its meaning significantly.

The doorway effect and sudden forgetfulness

We have all experienced the phenomenon of walking into a room and suddenly forgetting why we are there. Psychologists call this the "doorway effect." In these moments, we often mutter to ourselves, "Now, what was I doing?"

This usage is a form of cognitive retrieval. By vocalizing the question, we are trying to re-establish the mental context of our previous environment. The phrase here acts as a bridge between a lost intention and a current state of confusion. When users ask an AI this question, they are often experiencing a digital version of the doorway effect—flipping through tabs so quickly that they lose the thread of their original task.

Expressing regret or reflection in past narratives

Another powerful use of "what was I doing" is in the context of self-reflection or regret. When looking back at a poor decision, someone might ask, "What was I doing when I agreed to that?"

In this sense, the phrase implies that the speaker's past self was acting without full awareness, logic, or foresight. It is a way of distancing one's current, more informed self from the actions of a previous version of themselves. It suggests that the action was so out of character or illogical that it requires a retrospective investigation.

Comparing what I was doing and what I did

One of the most nuanced aspects of English is the choice between the past continuous ("I was doing") and the simple past ("I did"). The choice depends entirely on the "aspect" of the action—whether you see it as a completed point in time or a continuing process.

  • Simple Past (I did): This focuses on the completion of the task. "I did my chores." This means the chores are finished.
  • Past Continuous (I was doing): This focuses on the duration or the process. "I was doing my chores." This suggests that the focus is on the activity itself, often because something else happened in the middle of it.

If you say, "I know what I did," you are taking responsibility for a completed act. If you say, "I know what I was doing," you are often defending your intentions or the state of mind you were in while the act was in progress.

How to check your digital privacy and history

If you are concerned that your activities are being tracked, or if you genuinely need to find out what you were doing on your device, you should look at the tools specifically designed for that purpose, rather than asking a generative AI.

  1. Browser History: Every modern browser (Chrome, Safari, Firefox) has a "History" tab (usually accessed via Ctrl+H or Cmd+Y). This is the most accurate record of your web activity.
  2. Activity Monitors: On Windows, the "Task Manager" and on Mac, the "Activity Monitor" can show you what applications are currently consuming resources, though they are less helpful for looking at past history.
  3. Google My Activity: For those deeply integrated into the Google ecosystem, the "My Activity" page provides a centralized timeline of searches, YouTube views, and location history, provided these settings are turned on.
  4. AI Privacy Settings: Most AI platforms allow you to view and delete your prompt history. Checking these settings ensures that you are aware of what data is being used to "train" the models or "personalize" your experience.

Conclusion

The question "what i was doing" serves as a fascinating intersection between high-tech privacy concerns and everyday linguistic habits. Technically, AI assistants are designed with boundaries that prevent them from spying on your external activities, ensuring that your digital life outside the chat remains private. Linguistically, the phrase is a vital tool for navigating the complexities of time, memory, and self-reflection.

By understanding that an AI is not a surveillance device and mastering the grammar of the past continuous tense, users can interact with technology more confidently and communicate their experiences more clearly. Whether you are recovering from a "doorway effect" moment or reviewing your digital footprint, the key is to look in the right places and use the right language.

Summary

  • AI Access: Standard AI models cannot see your browser history, other apps, or your screen without explicit permission.
  • Context: AI only "knows" what you have shared in the current conversation.
  • Grammar: "What I was doing" uses the past continuous tense to describe ongoing past actions or provide background context.
  • Usage: The phrase is commonly used to express confusion, reflection, or to set the scene in a story.
  • Verification: To find your actual past activity, check your browser history or OS activity logs rather than asking an AI.

FAQ

Can an AI tell me what I was doing on my computer yesterday? No, a standard AI assistant has no access to your computer's system logs or historical data from other applications. It can only recall information from previous chats if you have history enabled and are in the same conversation thread.

Is it "What were I doing" or "What was I doing"? The grammatically correct form is "What was I doing." In English, the first-person singular pronoun "I" always takes "was" in the past tense.

Why does it feel like AI knows my interests if it isn't tracking me? This is often due to "data silos." While a conversational AI might not be tracking you, the search engine or social media platform you use might be. They use cookies to build a profile of your interests, which can sometimes make it feel like different technologies are "talking" to each other.

When should I use "What I was doing" instead of "What I did"? Use "What I was doing" when you want to emphasize that an action was in progress or to explain the circumstances surrounding another event. Use "What I did" when the action is finished and the completion is the most important part.

How can I protect my privacy when using AI? Review the privacy policy of the specific AI tool, use incognito or private modes if available, and avoid sharing sensitive personal information like passwords or financial data within the chat.