Building Custom GPTs for Public Affairs

Welcome to Inside Corporate Affairs, where we discuss the latest developments and best practices in achieving corporate affairs excellence. In this edition, is a US-China reset on the cards following Xi and Biden's meeting at the APEC Summit in San Francisco? David Cameron’s shock return, as we unpack Rishi Sunak's Cabinet reshuffle. And with 2024 set for a global election super cycle - how will this impact corporate affairs? All of this and more, with our focus of the week - building custom GPTs for public affairs.

This Week in Corporate Affairs

• US-China Reset as Xi and Biden Meet
• David Cameron’s Shock Return: Unpacking the UK Cabinet Reshuffle
• 2024: A Global Election Super Cycle 

US-China Reset as Xi and Biden Meet

Yesterday, Joe Biden and Xi Jinping convened in San Francisco at the APEC Summit to address pressing issues in US-China relations. The four-hour meeting yielded significant outcomes, including an agreement to reinstate military-to-military communications, which had been severed by China following a visit to Taiwan by then-House of Representatives Speaker Nancy Pelosi in August 2022. President Biden urged both countries to institutionalise these dialogues, with US Secretary of Defense Lloyd Austin set to meet his new Chinese counterpart in due course. This move is seen as an attempt to prevent misunderstandings and near-misses that have marked military interactions between the two nations over the past year.

Beyond the diplomatic achievements, this bilateral meeting holds implications for corporate affairs. A reduction in geopolitical tensions between the US and China could positively impact multinationals operating in both countries. Lowering the risk of trade disputes, sanctions, or regulatory challenges creates a more stable environment for businesses, fostering growth and investment. Moreover, the commitment to dialogue and cooperation on issues like artificial intelligence and climate change underscores the potential for collaboration in sectors with global significance. For organisations involved in these fields, the prospect of increased international cooperation could lead to expanded market opportunities and research collaboration. While challenges remain in this complex relationship, the meeting marks a step toward stabilising relations and addressing common global challenges. Corporate affairs leaders will no doubt monitor developments closely, hoping for a more predictable and cooperative environment that supports international engagement.

David Cameron’s Shock Return: Unpacking the UK Cabinet Reshuffle

In a significant Cabinet reshuffle orchestrated by Prime Minister Rishi Sunak, the United Kingdom witnessed notable changes in the government's composition. This reshuffle aimed to address the Conservative Party's declining popularity in public opinion polls and has sparked considerable attention and debate. A headline-making move in this reshuffle was the appointment of David Cameron, the former Prime Minister who resigned in 2016 following the Brexit referendum, as the new Foreign Secretary. Cameron's return to a prominent government role, especially after not being an MP since his resignation, came as a surprise. To facilitate this appointment, Sunak granted Cameron a life peerage, allowing him a place in the House of Lords. Beyond Cameron's appointment, the reshuffle included other significant changes. Home Secretary Suella Braverman, a prominent figure on the party's right wing, was dismissed, triggered by a dispute over the Metropolitan Police's handling of a pro-Palestinian march in London. James Cleverly, the former Foreign Secretary, replaced her as Home Secretary.

Corporate affairs leaders should stay vigilant, particularly as this Cabinet reshuffle is a response to the Conservative Party's declining popularity. As the 2024 election moves ever closer, the reshuffle reflects the government's bid to regain trust, but it may have upset internal party dynamics. As David Cameron assumes the role of Foreign Secretary, monitor shifts in foreign policy, especially regarding key players like China. Strike a balance between maintaining existing government relationships and building new ones.

2024: A Global Election Super Cycle 

2024 is gearing up to be a momentous period on the global political stage, marked by a surge in elections across 76 countries. According to the Economist, 43 of the countries are expected to conduct fully free and fair elections. Among them, 27 are EU members. However, a considerable portion of 2024's elections will occur in countries grappling with democratic deficits. Eight of the world's ten most populous nations are set for electoral showdowns. Yet, in half of these nations, notably Bangladesh, Mexico, Pakistan and Russia, elections face challenges such as limited freedom of speech and association, and are characterised by authoritarian rule. Conversely Brazil, India, Indonesia, and the United States fall under the ‘flawed democracies’ category in the Economist’s assessment, implying elections are generally free and fair but accompanied by systemic weaknesses.

For corporate affairs leaders, these elections bear significant consequences. Political stability, safeguarding democratic values, and predictable regulatory environments are critical to business continuity and investment decisions. In ‘flawed democracies,’ shifting political dynamics may introduce regulatory uncertainties or policy changes that businesses must endure.

Noteworthy elections in 2024 include Indonesia's pivotal legislative and presidential contests, India's struggle with anti-incumbency sentiment, and Brazil's municipal elections that could reshape the political landscape. Meanwhile the United States, with its presidential, congressional, and Senate elections in November, presents unique challenges. Renowned for its democratic traditions, concerns over political culture persist, making the outcome of US elections highly relevant for global business. Europe and Africa are also focal points, with the highest number of countries heading to the polls. These diverse elections will influence corporate affairs by shaping political and economic environments, particularly in Western Europe where migration issues dominate. Taiwan's January election, shaping its relations with China, holds particular importance for regional corporate interests – while Ukraine, amidst martial law and foreign occupation, is another exceptional case. A potential election, if held successfully, would signal resistance to Russia's influence but also holds risks for corporate interests.

How to Excel in Corporate Affairs

Corporate affairs leadership is critical for setting strategic direction, aligning engagement with business objectives and safeguarding an organisation's reputation. Effective leaders in corporate affairs navigate crises, shape public opinion, and foster a culture of integrity, ensuring that the function contributes strategically to their organisation’s overall success. Are you ready to take your corporate affairs leadership to the next level? If so my Corporate Affairs Senior Leader Coaching Programme may be a good fit for you.

This is a high-level coaching programme for global and regional heads of corporate affairs, public policy and related roles. Delivered remotely, and with the first cohort beginning in Q1 2024, you will develop your skills through one-to-one coaching with me in the six competence areas of 1) corporate affairs leadership 2) taking strategic action 3) powerbase development 4) influencing stakeholders 5) leading global teams and 6) managing risk and reputation. Together, these form the foundation of what it takes to excel in corporate affairs.

For one final week, this programme is being offered exclusively to Inside Corporate Affairs subscribers, but places are limited. To find out more and apply, contact me now at info@anordea.com.

Building Custom GPTs for Public Affairs

Artificial intelligence is reshaping the way we work, with large language models and generative AI disrupting and enhancing traditional workflows and processes. ChatGPT by OpenAI stands out as a leader in this space, and in a recent update, OpenAI have added the functionality to build custom GPTs. This represents the next step forward in the specialisation of GPTs, enabling more tailored and efficient use of language models to meet specific needs. In this article I’ll outline what custom GPTs are and the benefits they provide over standard models. I’ll also highlight how custom GPTs can be built using ChatGPT and explore how to optimise them for public affairs.

Introducing Custom Built GPTs

Custom GPTs represent a significant leap in AI technology, allowing for deep customisation to suit specific needs and contexts. At their core, GPTs are advanced language models capable of understanding and generating human-like text based on a wide range of inputs. Unlike standard GPT models, custom GPTs can be tailored to specific tasks, industries, or workflows. This customisation involves training or fine-tuning the model on specialised data sets, incorporating unique knowledge bases, and defining specific actions that the model can perform. Through this, it enables the creation of internal-only GPTs, built on top of an organisation’s knowledge base. In the context of public affairs, this means a GPT can be equipped to understand and interact with policy documents, stakeholder communication, media reports and more, making it an invaluable tool for professionals in the field.

The integration of custom GPTs in public affairs brings numerous benefits, making them an asset to public affairs teams.

• Enhanced Data Analysis: Custom GPTs can process and interpret vast amounts of information, from policy documents to public opinion trends, providing deep insights for strategic planning that integrates an organisation’s internal data.
• Efficient Communication: They can automate and personalise stakeholder communication, ensuring timely and relevant interactions that account for unique stakeholder profiles.
• Real-Time Monitoring: Custom GPTs can be used to monitor public sentiments and media trends through API integration, offering real-time insights critical for reputation management and strategic engagement.
• Predictive Capabilities: These AI models can predict public response to policies and media campaigns, more effectively than standard models, by learning from the results of past policy engagement.
• Content Creation: Custom GPTs assist in crafting communication materials, policy briefs, and press releases, ensuring consistency through alignment of the organisational context.
• Cost and Time Efficiency: By automating routine tasks, they reduce operational costs and save time, allowing public affairs professionals to focus on more strategic activities.

Overall, custom GPTs stand out as powerful tools in public affairs and all knowledge work, enhancing decision-making and operational efficiency. 

Building Custom GPTs

Building custom GPTs is user-friendly and doesn’t require extensive coding experience. The enables you to create your own GPT models simply through prompts, using a streamlined and intuitive interface.

Building a Basic GPT: OpenAI has provided a streamlined process for creating custom GPTs, accessible to Plus and Enterprise users.

• Access the GPT Editor: Go to the designated OpenAI link (chat.openai.com/gpts/editor) or select ‘My GPTs’ after logging in.
• Start Creating a GPT: Choose the ‘Create a GPT’ option.
• Use the GPT Builder: In the Create tab, communicate with the GPT Builder to specify what kind of GPT you want to create, like a GPT for assessing the impact of new policy proposals.
• Configure your GPT: Name and describe your GPT in the Configure tab, and select desired actions like web browsing or image creation.
• Publish your GPT: Once ready, you can publish your GPT and share it with others if desired.

Advanced Settings: OpenAI allow for further customisation with the GPT Editor. These advanced settings help you to get the most out of the customisation process.

• Add Images: Either request the GPT Builder to create an image or upload your own.
• Provide Detailed Instructions: Specify how the GPT should behave, its functionalities, and any behaviours to avoid.
• Prompt Starters: Set example prompts for starting conversations.
• Knowledge Resources: Provide context and additional information for the GPT to use.
• Enable New Capabilities: This includes web browsing, DALL·E image generation, and advanced data analysis.
• Custom Actions: Incorporate third-party APIs by detailing endpoints and parameters, and how the model should use them. You can also use existing plugin manifests to define actions for your GPT.

This process allows users to tailor GPT models to their specific needs, enhancing the usability and effectiveness of these models in diverse contexts.

Optimising the Building Process

When building custom GPT models, a structured approach to optimise the building process helps to ensure the effectiveness and relevance of your design, and improves its implementation across the public affairs function.

Defining Objectives and Targeting Specific Problems: Defining clear objectives ensures that the GPT's development aligns with specific goals, focusing efforts on achieving meaningful outcomes.

• Define Clear Objectives: Start by establishing clear goals for what your GPT is intended to achieve. This could be anything from improving policy monitoring to engaging with stakeholders and policymakers. Having well-defined objectives is key to developing an efficient and purpose-aligned GPT.
• Target Specific Problems: Identify the exact challenges or tasks your GPT will address. This helps keep the development focused and ensures the GPT is tailored to specific needs.
• Set Measurable Objectives: Establish quantifiable goals for your GPT, like reducing response times or increasing content output. This aids in evaluating the GPT's success post-implementation.

Data Quality and Ethical Considerations: High-quality data and ethical considerations are important to ensure the GPT's reliability, fairness, and compliance with data protection and ethical standards, building trust with users and stakeholders.

• Use High-Quality Data: Incorporate unique, high-quality data sources for more valuable AI applications.
• Prioritise Data Privacy and Ethics: Ensure your GPT complies with data protection laws and adheres to ethical standards. This includes being transparent about data usage and safeguarding against biases.
• Ensure Relevance and Effectiveness: The GPT should effectively address the identified needs. For instance, a GPT for monitoring policy should be fine-tuned to analyse data, while one for policy communications should facilitate narrative generation.

Integration and Compatibility: Focusing on integration and compatibility enables seamless incorporation of the GPT into existing workflows and systems, maximising its efficiency and utility.

• Workflow Integration: Consider how the GPT can be integrated into existing workflows to identify gaps where it could offer improvements, such as automating routine tasks or adding efficiency to daily routines.
• Focus on Integration and Compatibility: Assess how well the GPT can integrate with existing systems and tools already in use across the public affairs function.
• Plan for Scalability: Ensure that your GPT can scale with increasing demand and complexity.

Customisation and Continuous Improvement: Customisation for specific tasks and continuous improvement through testing and iteration enhance the GPT's adaptability and ability to meet evolving needs effectively.

• Customise for Specific Tasks: While GPTs are versatile, specialising them for specific tasks can greatly enhance their utility.
• Test and Iterate: Continuously test and improve your GPT based on user feedback and performance monitoring.
• Explore Advanced Features: Use features like API integrations and specialised knowledge databases to enhance your GPT's capabilities.

User-Centric Approach and Knowledge Updates: Understanding the user base and staying informed allow for a user-centric approach, while including examples helps to guide the GPT's behaviour, ensuring its relevance over time.

• Include Examples: Incorporate examples in the configuration of your custom GPT. For instance, a model built to evaluate new policy proposals might follow specific principles for text analysis.
• Understand the User Base: Tailor your GPT to its intended users and consider the needs and tech proficiency of your team.
• Stay Informed and Adapt: Keep up with the latest developments in AI and be ready to adapt your GPT to new advancements.

Altogether, by focusing on a structured approach to building GPTs, organisations can harness the power of AI-driven natural language processing to enhance workflows in public affairs.

Advanced Technique 1 – Knowledge Resources

Custom GPTs can leverage knowledge resources in their fine tuning and configuration, allowing users to provide extra context and information the GPT model can draw from when formulating a response. By uploading files, users can ensure their GPTs are informed by organisational-specific content that would otherwise not be included in the model's general training data.

The selection of knowledge resources will vary depending on the GPT model you chose to build and its intended purpose. However, a selection framework is likely to include a number of general features.

• Targeted Content Selection: Choose content that directly aligns with the intended application of the GPT model. This ensures that the GPT remains focused and relevant in its responses.
• Quality Over Quantity: Prioritise high-quality, accurate, and authoritative sources of information. The quality of the input directly affects the quality of the GPT's output.
• Diverse Sources: Include a range of perspectives and sources to avoid biases and ensure a well-rounded knowledge base.
• Scope and Depth: Aim for a balance between breadth and depth of knowledge resources to cover a range of topics, while also providing detailed information on specific subjects.
• Testing and Iteration: Regularly test the GPT's output for accuracy and relevance, and refine the knowledge base as needed.
• User Feedback Integration: Use feedback from users interacting with the GPT to refine and adjust the additional knowledge, tailoring it to real-world use and requirements.
• Regulatory Considerations: Be mindful of regulatory and privacy issues, especially when dealing with sensitive data or proprietary information.
• Regular Updates: Keep the additional knowledge up-to-date to reflect the latest information, especially in rapidly evolving fields.

For a public affairs team, knowledge resources could include a number of varying documents.

• Corporate Policy Documents: Details of the company's policies, ethics, and compliance guidelines.
• Press Releases Archive: A collection of past press releases for reference on company messaging and history.
• Stakeholder Communications: Examples of communications with key stakeholders, including policymakers, key trade associations and coalition partners.
• Industry Reports: Information on industry trends, challenges, and benchmarks.
• Crisis Management Plans: Documents outlining procedures and communication strategies for potential crises.
• Legal and Regulatory Updates: Information on relevant legal and regulatory changes impacting the industry.
• ESG Initiatives: Details of the company's ESG activities and sustainability reports.
• Media Coverage Database: A compilation of past media coverage for understanding public perception and media relations history.
• Internal FAQs and Knowledge Base: Commonly asked questions and answers about the company, its mission, goals and objectives.
• Speeches and Presentations: Archive of speeches and presentations by senior leaders for tone and messaging consistency.

These elements would significantly enhance a custom GPT's ability to provide relevant, accurate, and context-aware responses in public affairs. Altogether, when building your custom GPT, it's essential to carefully consider the type of knowledge and data sources you integrate. This ensures your GPT is not only tailored to specific tasks but also capable of drawing on relevant, accurate, and up-to-date information to enhance its responses and functionality.

Advanced Technique 2 – Custom Actions

Custom actions refer to specific, user-defined functionalities that an AI is configured to perform. Unlike standard actions, which are generic and built into the model, custom actions are tailored to meet specific needs or objectives. This makes the model more adaptable to niche or complex tasks.

Custom actions provides a number of benefits. They allow for the creation of a GPT model that is highly tailored to specific industries, organisations, or even individual projects, providing more relevant and targeted outputs. By automating specialised tasks, custom actions can significantly reduce manual effort and time spent on routine or complex tasks, leading to greater operational efficiency. Custom actions are designed with particular objectives in mind, which can lead to higher accuracy and effectiveness in those areas compared to using a standard GPT model. As needs evolve, custom actions can also be modified, offering scalability and flexibility to adapt to changing requirements.

Zapier, a platform known for automating workflows by connecting different apps and services, can play a significant role in integrating custom built GPT models into broader workflows and systems. By using Zapier, you can connect your custom GPT model with hundreds of other applications and services, enabling seamless data exchange and triggering actions across different tools. For instance, you could set up an automated workflow where the output of your custom GPT model triggers actions in other apps, like sending emails, updating databases, posting to social media, or creating tasks in project management tools. This integration helps in harnessing the full potential of custom GPT models by embedding them into complex workflows.

Value Chains and Workflows in Public Affairs

Perhaps the biggest challenge in building custom GPTs is developing use cases in public affairs. This goes beyond what OpenAI’s GPT Builder can assist with, instead requiring a deep understanding of how value is generated in public affairs and how this value maps to activities and workflows.

Value chains are a well-established concept in business and strategic management, primarily used to analyse and optimise the sequence of activities that an organisation undertakes to create value. The concept of value chains has proven highly effective in enabling organisations to identify opportunities for cost reduction, process improvement, and differentiation within their industry.

Public affairs, often regarded as a multidisciplinary function within organisations, plays a pivotal role in managing relationships with external stakeholders, influencing public policy, and safeguarding an organisation's reputation. This complexity, makes it important to have a structured approach to workflow management. Incorporating the concept of value chains into public affairs not only aids in understanding and optimising workflows, but also aligns the operations of the function with best practices. Leaders can therefore leverage value chain principles to streamline processes and achieve goals more effectively.

Mapping workflows within a public affairs function using a value chain framework provides a systematic and structured approach. Workflow mapping involves the visual representation and documentation of the various steps, tasks, and processes that comprise a specific workflow or business process. This mapping allows for a clear and comprehensive understanding of how activities are carried out, who is responsible for each task, and how they relate to one another. In the context of public affairs, workflow mapping provides a blueprint for managing complex interactions with stakeholders, policymakers, and the media. Altogether, process mapping involves several discreet steps.

• Identifying Primary and Support Activities: Begin by categorising the primary and support activities that make up the public affairs function. Primary activities often involve strategic communications and stakeholder engagement, while support activities may include strategic alignment, and resource allocation.
• Analysing Dependencies and Interactions: Understand how these activities are interrelated and dependent on one another. Recognise where one activity feeds into another and where dependencies exist.
• Process Mapping of Key Activities: Create detailed process maps or flowcharts for each key activity within the value chain. These maps should outline the sequence of tasks, responsible individuals, and the flow of information.
• Assessing Efficiency and Cost Effectiveness: Evaluate the efficiency of each workflow by identifying areas of delay, redundancy, or inefficiency. Consider the associated costs and potential cost-saving opportunities.
• Quality and Performance Metrics: Define KPIs and quality metrics for each workflow to measure their effectiveness and impact.
• Identifying Technology and Automation Opportunities: Explore how technology and automation can be integrated into workflows to streamline processes and reduce manual efforts. This of course is where GPTs fit in.
• Employee Input and Collaboration: Involve employees who are directly involved in these workflows to gather insights, feedback, and suggestions for improvement. Collaboration fosters a sense of ownership and commitment to the optimised workflows.
• Continuous Improvement Strategies: Develop a plan for continuous improvement, implementing changes gradually, monitoring their impact, and making further adjustments as needed to optimise workflows over time.

By following these steps, public affairs teams can harness the power of value chains to map, optimise, and enhance their workflows, ultimately enabling the successful integration of custom built GPTs.

A Portfolio Approach to Custom GPTs

While a single, general-purpose GPT model can perform impressively in a wide range of tasks, it may not excel in any particular domain. This can lead to suboptimal results when addressing specific, specialised challenges. A focus on workflow allows for the adoption of not just one custom built GPT, but a portfolio of GPTs. This approach involves the creation and utilisation of multiple GPTs, each specialised for specific tasks.

This approach has several benefits including improved task specificity, tailored solutions for varied workflows, and enhanced robustness.

• Specialisation: Each GPT within a portfolio can be tailored to excel in a specific area of expertise. One GPT may be optimised for policy analysis, while another focuses on policy communications. This specialisation enhances the accuracy and relevance of responses.
• Precision: Specialised GPTs tend to generate more precise and context-aware outputs, making them invaluable for tasks that demand a high degree of accuracy.
• Workflow Alignment: Organisations can align GPT applications with various departments by customising GPTs for their specific use cases. This alignment optimises the impact of AI across the organisation.
• Adaptability: As organisational priorities evolve, a portfolio can adapt accordingly. New GPT models can be developed or existing ones fine-tuned to address emerging challenges, ensuring ongoing relevance and effectiveness.
• Risk Mitigation: By diversifying GPT models in a portfolio, organisations can mitigate the risks associated with model failures or biases. If one model encounters challenges or exhibits biases in its responses, others can continue to operate effectively.

By adopting this approach, organisations can effectively implement and manage a portfolio of custom GPTs, thereby capitalising on the full suite of advantages that custom GPTs provide.

Conclusion

The integration of custom GPTs in public affairs signifies a major step forward in how policy analysis, strategic communication and stakeholder engagement are managed. These advanced AI tools offer unprecedented capabilities in data analysis and campaign design. While OpenAI is a leader in this space, public affairs professionals should keep up-to-date with changing industry dynamics and capitalise on fresh opportunities to harness the power of AI for strategic advantage.

Leadership Takeaways

• Custom GPTs represent a significant advancement in AI technology, allowing for deep customisation to suit specific needs and contexts.
• Custom GPTs can be tailored to specific tasks, industries, or workflows, making them invaluable tools for professionals in various fields, including public affairs.
• The integration of custom GPTs in public affairs brings benefits like enhanced data analysis, efficient communication, real-time monitoring, predictive capabilities and content creation.
• Building custom GPTs is user-friendly and doesn't require extensive coding experience, making it accessible to a wide range of users.
• Advanced settings in the GPT Editor allow for further customisation, including adding images, providing detailed instructions, setting prompt starters, and enabling new capabilities.
• Knowledge resources play a crucial role in fine-tuning and configuring custom GPTs, enabling them to provide context-aware and informed responses.
• Custom actions are specific functionalities that can be configured for AI models, enhancing their adaptability and effectiveness in niche or complex tasks.
• Integrating custom GPT models into workflows using platforms like Zapier can streamline data exchange and automate actions across different tools.
• Mapping workflows in public affairs using a value chain framework provides a structured approach to optimising processes and integrating GPTs effectively.
• Defining clear objectives and targeting specific problems is essential when building custom GPTs to ensure alignment with goals and meaningful outcomes.
• Prioritising high-quality data and ethical considerations ensures the reliability, fairness, and compliance of custom GPTs.
• Focusing on integration and compatibility enables seamless incorporation of GPTs into existing workflows and systems, maximising efficiency and utility.
• Continuous improvement through testing, iteration, and staying informed about AI developments enhances the adaptability and relevance of custom GPTs.
• A portfolio approach to custom GPTs, specialising each model for specific tasks, enhances task specificity, precision, adaptability, and risk mitigation.  

That's it for this week's edition of Inside Corporate Affairs. Subscribe now, and if you like what you read today, please like and share it with your network to help me reach a wider audience. Have a good day, a great week, and I'll see you again soon.