Jan. 2026 Magazine - Flipbook - Page 16
TECH TALK
Best Practices for Surveyors Using AI
Maintain Professional Responsibility: AI is a tool, not a replacement for professional judgment.
Always verify AI-generated or revised writing against primary sources, regulations, and your
professional expertise. Never rely solely on LLM outputs for final boundary determinations or
stamped professional work.
Engage Your Critical Thinking First: Whether researching regulations or drafting
correspondence, start by thinking through what you need to know or communicate. Write
down your key points, questions, or requirements. Continuous engagement of your
professional knowledge makes your prompts more effective and keeps your critical thinking
skills sharp.
Be Specific and Contextual: Include relevant context in your prompts and key source materials.
Mention the persona the LLM should take on (surveyor), jurisdiction, applicable standards, and
specific technical requirements. The more specific your prompt, the more useful the response.
Iterate and Refine: Like adjusting an instrument or mentoring a junior surveyor, refining your
prompts based on initial results improves outcomes. If the first response is not quite right,
clarify or add detail. This iterative process is part of effective prompt engineering.
Understand Limitations: Some AI systems have knowledge cutoff dates and cannot access
case-specific field conditions. They may miss nuances or provide incomplete information that
requires your professional expertise to identify and correct.
Protect Client Confidentiality: Never input confidential client information, specific property
details, or sensitive project data into cloud-based AI systems. The information added to a
commercial LLM may be used to refine and train the system. For surveyors who need to
process sensitive project data with AI assistance, locally-installed LLMs offer a secure
alternative. These local systems run entirely on your own hardware without transmitting data to
external servers, allowing you to maintain confidentiality while still leveraging AI capabilities.
Choosing and Committing to an LLM
Several Large Language Models (LLMs) are available to professionals, including ChatGPT, Claude,
Gemini, and others. We suggest experimenting with different platforms initially to determine which
best suits your communication style and professional needs. Try the same prompts across different
LLMs and evaluate which produces outputs that align most closely with your expectations and
workflow. Each system has different strengths, interface designs, and response patterns.
For those with data security concerns or working with sensitive project information, local LLMs
offer an alternative worth considering. Tools like Ollama allow you to run AI models directly on
your own computer or server, and interfaces like OpenWebUI provide a familiar ChatGPT-like
experience for interacting with these local models. Because local LLMs run entirely on your
hardware without sending data to external servers, they provide enhanced security and privacy.
Additionally, local installations have no usage limits, making them cost-effective for high-volume
tasks. However, the trade-off is reduced processing power compared to commercial cloud-based
systems. Local LLMs typically run slower and may not handle complex tasks as deeply as their
commercial counterparts. For everyday tasks like drafting routine correspondence, organizing
notes, or generating basic reports, local LLMs can perform quite well.
Once you identify the LLM approach that works best for your practice, commit to using it
consistently. This allows the tool to better adapt to your communication patterns and preferences
over time, since many now incorporate memory features that learn from your interactions.
Looking Forward
Prompt engineering represents a shift in how we interact with technology. Rather than learning
complex software interfaces, our latest challenge is learning to communicate effectively with LLMs.
The surveying profession has always balanced traditional knowledge with new technology. Just as
we teach fundamentals not merely to preserve old methods but to train the minds of young
surveyors toward professional problem-solving, we must approach LLMs as tools that enhance
rather than replace our own skills. As with any tool, its value depends on the skill and professional
judgment of the person using it.
The question is not whether AI will become part of surveying practice, but instead how we as
professionals will integrate it responsibly and effectively into our work. Learning the skill of prompt
engineering now, while maintaining our commitment to critical thinking and professional judgment,
positions us to use AI wisely while upholding the high standards the profession demands.
14 January 2026 | THE TEXAS SURVEYOR
