Hey ChatGPT, can you build me an estimate?

“Artificial intelligence is a machine’s ability to perform the cognitive functions we usually associate with human minds”

— https://www.mckinsey.com/featured-insights/mckinsey-explainers/what-is-ai

Is AI software in cost estimating poised to be an industry game changer, much like the Tesla Model S which revolutionised the automotive industry? As much as the majority would love to take a Tesla for a burn (mind the pun) not all are sold on AI and its implied value, so let’s explore it.

Artificial Intelligence is used across a multitude of formats and technologies. In its basic form, AI supports all of us in our everyday pursuits, such as:

• Personalising online experiences – tailoring advertising and providing content recommendations based on user behaviours and preferences

• Editing content – autocorrecting and suggesting improvements in written content across various platforms

• Strengthening security and identification – using facial recognition and fingerprint identification to enhance security measures

• Automating administrative tasks – reducing administrative workloads and overheads through digital assistants and chatbots,

• Educational automation like ChatGPT and similar technologies that use deep learning to teach and answer queries, reducing the human workload in information sharing and education.

Will AI make its way into activities like cost estimating?

Traditional cost estimating methods such as First Principles and Value Engineering have withstood disruptive technology for years, still, a large portion of the workforce use Microsoft Excel to develop budgets and pricing models.

Others embraced estimating applications such as Expert Estimation and Candy Estimating, which optimise cost models, and streamline the estimating process, enhancing the accuracy and consistency of project budgets.

Whether you use Microsoft Excel or Expert Estimation, the Estimator must have a sharp eye for detail, strong analytical skills, excellent communication, industry knowledge, ethical integrity, and most importantly, a significant level of patience.

AI powered estimating tools, do they stack up?

Using AI powered applications like ChatGPT for cost estimating, appears promising given its ability to analyse big data such as historical costs, market trends, project specifications, and environmental factors to generate estimates using complex algorithms and models.

Sounds impressive and it is, but like the new Tesla S owner with the number plate “L0LG4S”, I remain conflicted, will AI software really take over the role of a Cost Estimator?

There are some broad risks that come to mind when pondering the efficacy of AI produced cost estimates, such as:

• Ethical and bias concerns

  The potential for learning and perpetuating biases in training data, could produce discriminatory estimates presenting key ethical issues.

• Accuracy and reliability

  AI dependency on high quality training data, presents potential risk of inaccurate data, resulting in unreliable estimates.

• Loss of human expertise

  Relying on AI poses the obvious risk of declining human expertise, presenting a potential skill gap when human input is required.

• Security and data privacy

  Exposing sensitive data to potential unauthorised access and privacy breaches.

• Intelligible and transparent

  The complex nature of AI algorithms make it hard to understand and explain how estimates are derived, posing issues in contractual contexts.

Artificial Intelligence analyses data faster than the human, generating detailed reports that, on the surface, appear to grasp concepts, and manage changes on the go, adjusting estimates based on perceived market conditions, labour costs, material prices, and more.

“The volume and complexity of data that is now being generated, is too vast for humans to reasonably reckon with and has increased the potential of machine learning, as well as the need for it.”

— https://www.mckinsey.com/featured-insights/mckinsey-explainers/what-is-ai

Whilst I support dramatically reducing manual effort and accelerating productivity, believing that AI will surpass the human element of cost estimating has some flaws.

Hold onto your calculators, trusty (often corrupt) spreadsheets and proven methods… at the end of the day, AI is a tool, just like Expert Estimation and others, the output quality is only as good as the data input and the people analysing it.

I invite my fellow professionals to weigh in: How do you think AI will shape the future of cost estimating?

Written by Scott Larson, Senior Project Engineer/ Estimator at JC Civil Projects 

Scott is a Registered Professional Engineer of Queensland civil and mining engineer with over 12 years’ experience working in both onsite and offsite engineering roles within the civil construction industry, delivering first principle estimates through to and project managing infrastructure upgrades, mine rehabilitation, large master-planned subdivisions, educational and community facilities, water infrastructure, major road transport upgrades, mining and renewable energy projects. 

Drawing on his experience in the pre-contract and project delivery phase to identify cost efficiencies and advise on constructability, Scott delivers cut-through value propositions and estimates that have seen him win tenders valued in excess of $30m. A lifelong learner, Scott immerses himself in new technologies, tools, programs, methodologies and practices, with a focus on identifying the technologies that will drive innovation and re-shape the engineering industry.