From Six Months to Six Seconds: How AI Turned Me into a Bureaucratic Superhero (And Saved My Sanity)
The Impossible Dream: Government Plans at the Click of a Button
Imagine a world where creating any report or document, in this case a council plan, is as easy as clicking a button. No more months of gruelling effort, countless meetings, or tedious data crunching. Sounds too good to be true, right? Well, thanks to the power of Artificial Intelligence, this is no longer a distant dream but a tangible reality. In this weeks “what I have been doing” piece, I’ll take you through my experience of harnessing AI to revolutionize a use case for the way we develop council plans. Whether you’re a tech professional, a government official, or an AI enthusiast, I hope to show the way and inspire transformative potential of AI in streamlining local government operations and bringing about unprecedented efficiency and innovation.
Creating council plans has always been difficult, really difficult, involving extensive coordination, numerous staff members, and heavy reliance on qualitative data, multiple levels of engagement inside the organisation as well as community engagement. As a consultant working with various government bodies, I witnessed firsthand the challenges and inefficiencies of this traditional approach. It was clear to me that there had to be a better way, one that leveraged modern technology to simplify and enhance the process.
That’s when I embarked on building my AI tool: using AI to develop council plans. My goal was simple yet ambitious — to replace the labour-intensive, time-consuming methods with a more scientific, data-driven approach. By ingesting vast amounts of data from existing council plans, census records, and community inputs, I aimed to create a system that could generate high-quality council plans in a fraction of the time.
The results were very positive and can demonstrate how to make government at all levels potentially “effective”. What previously took six months could now be achieved in just two minutes. The cost savings were significant, freeing up resources for other important initiatives. But the journey wasn’t without its challenges. From skepticism among councillors to concerns about information security, there were hurdles to overcome. Yet, the potential benefits far outweighed the obstacles, paving the way for a new era of efficiency and innovation in local government planning.
This week thought piece I delve deeper into this, sharing insights, discoveries, and the transformative impact of AI on something innocuous and simple as a council plan.
Confessions of a Government Consultant: My AI Adventure
I have a lot of experience as a consultant for different government agencies, where I helped them use technology to enhance their work, but if I put my personal view forward, I doubt consulting firms have really achieved the impact they say they did. Recently, I embarked on an exciting personal project: using AI to develop council plans. The traditional method of creating these plans is labor-intensive, requiring the coordination of numerous staff members and relying heavily on qualitative data and low sample data of community needs. I knew there had to be a better, more data driven way.
With AI I started with the simple idea that data, when harnessed correctly, could provide a more scientific and efficient approach to planning. Over time, I’ve developed several AI tools, but this particular project of creating council plans has been eye opening and rewarding this year.
To view the output I have supplied a link below:
Bureaucracy Busting: How AI Took on the Council Plan Challenge
When choosing a proof of concept, it’s crucial to identify a process that is ripe for improvement. In this case, creating council plans stood out. The traditional process involves many staff members, extensive organizational engagement, and is based mostly on qualitative data rather than quantitative insights. This not only makes the process lengthy but also subjective and prone to inconsistencies.
By focusing on this use case, I aimed to demonstrate how AI could transform a cumbersome task into an efficient, data-driven process. The time and effort to create a council plan were significant, with multiple departments involved and a heavy reliance on manual data collection and analysis. It was the perfect candidate for AI intervention, offering the potential to save time, reduce costs, and improve the overall quality and consistency of the plans.
Feeding the Beast: 75 Council Plans, 150MB of Data, and One Very Hungry AI
To begin the process, I needed a substantial amount of data. I ingested 75 council plans from various local governments across Australia. This provided a rich dataset that captured the diversity and commonalities in council planning processes and structure. Additionally, I gathered 150MB of Excel files containing census data, demographic information, and inputs from community channels Australia wide. This combination of sources ensured that the AI had a comprehensive understanding of both qualitative and quantitative factors influencing council plans.
The real magic happened when I fed this data into the AI system. The AI used natural language processing (NLP) to analyze the text of the council plans and extract key themes and objectives. It then cross-referenced this information with the quantitative data to identify trends and correlations. This allowed the AI to generate draft council plans that were not only based on historical data but also tailored to current demographic and community needs.
Déjà Vu: When All Council Plans Start to Look the Same
One of the most striking discoveries was the lack of variation in council plans. Despite the different communities and local needs, many plans followed a similar structure and set of objectives. This groupthink mindset indicated that councils were often using a templated approach, which limited the uniqueness and effectiveness of their strategies.
The AI’s ability to analyze multiple plans simultaneously highlighted these patterns. It became evident that while each council faced unique challenges, their strategic objectives were often remarkably similar. This revelation underscored the potential for AI to introduce more innovation and customization into the planning process. By breaking away from the traditional template, councils could develop plans that better reflected their specific communities and needs.
From Gut Feelings to Hard Facts: AI’s Quantitative Takeover
The real breakthrough came when we overlaid quantitative data onto the qualitative insights from the council plans. This data-driven approach allowed us to match statistical information with narrative content, providing a more holistic view of the needs and priorities of each community. For instance, census data on population growth, age distribution, and economic indicators could be directly linked to strategic objectives in the council plans.
This integration of data not only made the plans more robust but also more actionable. Quantitative metrics provided clear benchmarks and performance indicators, enabling councils to track their progress more effectively. The AI could also identify gaps and suggest areas for improvement based on empirical evidence, making the planning process more dynamic and responsive to changing conditions.
Overall, the combination of qualitative and quantitative data transformed the council planning process. What used to take months of manual effort could now be achieved in a matter of minutes, with a level of precision and insight that was previously unattainable.
Time Travel is Real: Turning Six Months of Work into Six Seconds
The most immediate and impressive result of using AI to create council plans was the drastic reduction in time required. What once took six months, or more, of painstaking effort could now be completed in just two minutes. This efficiency gain wasn’t just about saving time; it translated into significant cost savings as well.
Traditionally, creating a council plan involved numerous staff members, extensive engagement sessions, and countless hours of analysis and writing. The direct and indirect staff hours alone were estimated to cost around $250,000. When factoring in the efforts of executives, councilors, and external consultants, the total cost of producing a single council plan could reach up to $375,000. By automating this process with AI, these costs were slashed to near zero, freeing up resources for other important initiatives or providing councils with the option of 3 new hires.
Fear and Loathing in the Council Chambers: The AI Backlash
Despite the clear benefits, the introduction of AI in creating council plans wasn’t without its challenges and criticisms. One of the primary sources of negative feedback came from councilors themselves. Many were apprehensive about a technology they didn’t fully understand and feared the implications of automating such an integral part of their governance.
There were also valid concerns about information security. Since the AI relied on publicly available data posted on council websites, it raised questions about the potential misuse of this information. The ability to ingest, train, and use public data for new purposes posed a threat not just to local governments but potentially to national security as well. If we use this data to train models the way I have we rely on the fact that the information available is gold standard. What I may have train my model on is imperfect documents, cleansed of political risk and standardised. This may be the biggest flaw in AI for government use cases.
Moreover, the infallibility of AI was another major point of contention. AI systems can and do make errors, and there needs to be a robust mechanism to challenge and verify the outputs they produce. This is particularly important when the outcomes of these AI-generated plans directly impact communities and people’s lives.
Local Government in a Box: Why the Future is Both Exciting and Slightly Terrifying
The reality, from my experience, is that many council plans end up as “set and forget” documents rather than dynamic guides that local governments follow religiously. While the risks associated with AI-generated plans are real, they are often mitigated by the fact that these plans are not always actively implemented. However, this doesn’t mean we should become complacent.
Looking forward, I can envision a future where we have “local government in a box” — a system where a significant volume of local government activities are automated. This could lead to greater efficiency and allow councils to move away from reports and low value works and put more staff into community facing work to truly improve communities. To achieve this, it’s essential to focus on use cases that consume significant organizational effort and look at the potential for hours saved rather than just financial figures.
Another crucial step is to conduct an update lens and thorough security reviews of the information published on council websites. As AI becomes more prevalent, the data we put out publicly will be used to train various models. Ensuring this data is secure and appropriately managed and accurate is vital to safeguarding our communities and maintaining trust in the systems we use.
Embracing the AI Revolution: The Future of Council Planning is Here
AI has the potential to revolutionize the way we create and implement council plans. My experience has shown that with the right approach, we can turn a labor-intensive process into an efficient, data-driven system that saves time and resources while providing more robust and actionable plans.
As we continue to explore and embrace AI in government, it’s important to address the concerns and challenges head-on. By doing so, we can unlock the full potential of AI to drive innovation, improve governance, and better serve our communities. Let’s not shy away from the future — let’s shape it with the power of AI.
Enjoying my articles?
1. Like and comment — I try to reply to everyone.
2. Follow Me Here or on LinkedIn
