As artificial intelligence chatbots become ubiquitous tools in our daily lives, a critical skill is emerging: the art of crafting the perfect prompt. According to Amanda Askell, a philosopher at leading AI company Anthropic, most users are likely approaching this task incorrectly. In a recent Q&A video released by the company, Askell dismantled the idea of a one-size-fits-all manual, framing prompt writing as a dynamic, experimental practice essential for unlocking an AI's full potential.
There's No Textbook for Talking to AI
Amanda Askell, whose philosophical training informs her work, described prompt engineering as an "empirical domain." This means there is no single rulebook. Instead, users must learn through observation and iterative testing. "Prompting is very experimental," Askell explained. She shared that her approach changes with each new model, developing a unique strategy through extensive interaction.
The key, she argues, is to discard preconceived notions. Users need to patiently review "output after output" to understand the specific tendencies and disposition of the model they are using. "It is really hard to distill what is going on because one thing is just like a willingness to interact with the models a lot," she noted. This process of trial, error, and careful analysis is fundamental.
Clarity and Context: The Philosopher's Approach
Askell believes her background in philosophy is surprisingly relevant. "This is where I actually do think philosophy can actually be useful for prompting," she stated. The core of her method involves translating complex thoughts into clear, explicit instructions for the AI. Much of her job involves meticulously explaining an issue, concern, or idea to the model with as much precision as possible.
This aligns with official guidance from Anthropic. In a 'Prompt Engineering Overview' published in July, the company offered a powerful analogy for users. It advised thinking of its chatbot, Claude, not as mere software but as "a brilliant, but very new employee (with amnesia) who needs explicit instructions."
Anthropic's Blueprint for Better AI Conversations
The guide underscores a crucial limitation of current AI: it lacks persistent memory of past interactions. Unlike a human colleague, Claude has no inherent understanding of "your norms, styles, guidelines, or preferred ways of working." Every conversation is a fresh start. Consequently, the quality of the output is directly tied to the quality of the input.
"The more precisely you explain what you want, the better Claude's response will be," Anthropic noted. The company emphasized that providing rich contextual information is paramount. "Just like you might be able to better perform on a task if you knew more context, Claude will perform better if it has more contextual information," the overview stated.
This expert insight arrives at a pivotal moment. With AI integration accelerating across education, business, and creative fields, moving beyond basic commands is no longer optional. Mastering this new language—learning to 'reason' with AI by providing structured, clear, and context-rich prompts—is becoming a fundamental digital literacy skill for users in India and worldwide.
