Cleve Gibbon

content management, content modelling, digital ecosystems, technology evangelist.

Value preparation over generation

Currently, there is significant emphasis on using Generative AI (GenAI) to create more content. Beyond the well-documented legal, ethical, and governance risks associated with GenAI, prioritizing value preparation over generation serves as an overarching mitigation strategy for delivering future-ready outputs.

Being pragmatic and practical can be frustrating, and it may seem to constrain innovation during a period of massive, rapid technological disruption. However, having the right data to train your models to achieve optimal outputs is strategically your best next move. This isn’t merely a choice between low risk or high innovation, but rather a balance where considered risk can unlock innovation. You need a combination of both risk and innovation to make progress.

Value preparation over generation

Value Preparation over Generation

In the rush to generate, many are shortcutting preparation. This approach may yield short-term gains but fuels long-term pain. When decorating a room, for example, it’s about 95% preparation and 5% painting. Stripping back to the plaster, smoothing out cracks, sanding down the woodwork, and cleaning debris are time-consuming but necessary preparation tasks. The quality of the final coat depends heavily on the quality of these preparatory steps.

GenAI is no different. The various hoops you must jump through to prepare data for training are critical to ensuring the accuracy and quality of downstream generative outputs. It’s the small things that count. For instance, we discovered that when preparing assets to train a brand-specific custom model, centering the images within the digital asset was crucial. The more training images required, the more manual effort needed during the preparation phase. This quickly becomes a hurdle that automation can help overcome—and we are making strides in this area. However, this is just one of many tasks being added to a growing pipeline of preparation tasks.

In Summary

The importance of data preparation in training AI models is a burgeoning area of research. Invest time to take the long way around. Only then can you creatively, iteratively, and incrementally shorten the distance between your inputs and desired outputs.

Slow Productivity

Slow Productivity

I’ve always had a keen interest in productivity. Starting out with getting things done (GTD) and also plans, progress, and problems (PPP) reports. And then I just happened across Slow Productivity by Cal Newport.

Cal talks about pseudo-productivity that is basically a focus on busyness. I learnt from an early age that busy is failure to prioritze. Slow productivity is about how to focus and execute on the right things by following three simple principles:

  1. Do fewer things
  2. Work at your own pace
  3. Obsess over quality

The books gives many examples of key people that succeeded using slow productivity. From Isaac Newton to Alanis Morrisette (which I’m looking forward to seeing this August 2024).

If you’re an overworked knowledge worker it’s worth a read.


AI Agent Builders

Google recently unveiled Vertex AI Agent Builder. This new tool allows for the creation of AI assistants. Despite hopes for a zero-code approach, technology proficiency remains essential. And just like Amazon, Microsoft, and IBM who are targeting enterprise users, Google’s Vertex needs experienced technical users to micromanage AI from design through to deployment.

AI Agent Builder

At the other extreme, something like launch lemonade is an AI agent builder for the everyday user.  Good enough to get something up and running, monetizing AI agents from the get-go. However, not powerful enough yet to develop truly differentiated, enterprise grade, products. 

Then you have those platforms that live somewhere between these two extremes. Rasa specializes in creative conversational AI agents with a strong emphasis on natural language understanding. While DataRobot is a platform embraced by data super heavyweights to enable users to create deployable autonomous predicative models.  And then there is the platform of experts, or POE.  This platform creates AI agents capable of decision making and task execution. 

In short, it’s a mixed bag but the message is clear.  AI Agents are on the rise.  I’m waiting to see just how human friendly GPT5 AI Agents are.

From AI Agents to AI Companions

As a teenager, I devoured ‘The Culture’ series by Iain M. Banks, which started in 1986. The series portrays ‘The Culture’ as a society that has moved beyond scarcity, where AIs play a crucial role in governance and societal structure. That’s right, many AI agents. Minds and Subminds possessing vast computational capabilities, personality and autonomy. Drones that service as assets, workers, and companions. And finally avatars that were physical extensions of other AIs us to interact with more directly with with biological beings. Banks was ahead of his time, exploring the co-existence of AI and humans.

AI Agents to AI Companions

The narrative of AI agents as human partners is still unfolding. Today AI agents are clunky, task-based, and largely confined to the realm of the tech-savvy. In the corporate world, AI agents, or co-pilots, are increasingly augmenting the digital workforce. Our AI agents summarize meetings, write blogs post, perform customer service, and make recommendations. But can they confidently unload your inbox. I don’t think so. That requires a higher level of sophistication, tact, planning, and intelligence.

However, in our domestic lives, we desire AI companions that help us accomplish more. And I prefer the term ‘companion’ over the corporate terms like ‘agents,’ ‘partners,’ and ‘co-pilots’. I want the trust of a companion with whom I can do things on a deeply personal and creative level. I wouldn’t share my companion, and my companion wouldn’t be suited to anyone else. My companion is an extension of myself. My AI companion could organize my wife’s 40th birthday party for me. However, my AI agent would make recommendations for venues with enough prompting. We aren’t there yet with AI companions, but that’s the goal.

So, add The Culture series to your reading list for a glimpse into the near future.

ai agents in the culture series

The Culture Series

  1. Consider Phlebas (1987) – The first published novel of the series, set during the Idiran-Culture War.
  2. The Player of Games (1988) – Follows a Culture citizen who is an expert game player that is recruited by the Culture.
  3. Use of Weapons (1990) – Centers on an operative in the Special Circumstances division of the Culture.
  4. The State of the Art (1991) – A collection of short stories and a novella, with the title story dealing directly with the Culture.
  5. Excession (1996) – Involves the Culture’s encounter with an enigmatic and powerful artifact known as the Excession.
  6. Inversions (1998) – A novel that can be read as a Culture book or as a standalone story, featuring two parallel stories that may involve Culture agents.
  7. Look to Windward (2000) – Set in the aftermath of the Idiran-Culture War, focusing on the effects of the war on different individuals.
  8. Matter (2008) – Explores the interactions of advanced and primitive societies within the Culture’s universe.
  9. Surface Detail (2010) – Deals with the ethics of simulated realities and the afterlife.
  10. The Hydrogen Sonata (2012) – The final novel published before Banks’ death, concerning a civilization preparing to Sublime, a concept frequently mentioned in the series.

The Prompt Recap

the prompt recap

Creating effective generative prompts is a skill you can master with practice (openai, claude, etc). Here’s an additional tip you should use to refine your technique for better results. I call it the prompt recap. So:

  1. Start a conversation: Prompting is not a one-off; it’s an interactive exchange. Begin by asking for what you need.
  2. Evaluate the response: Look at what the AI produces. Is it close to what you had in mind?
  3. Refine your request: Make adjustments based on the response. Sometimes it’s the small tweaks that bring big improvement until you have your desired output.
  4. Do the prompt recap: But don’t stop there. Once you have the desired result, ask the AI for a prompt recap: can you replay the prompt to generate this output? It will then produce an all-in-one prompt that gets you to your desired out.

The prompt recap shows you how the AI would construct the prompt. In doing so, it’s teaching you the ‘why’ behind the ‘what’. And from this you gain insights into how to craft prompts that get straight to the desired outputs.

Give it a go and let me know how you get on!

Towards AI Agents

The digital workforce is on the cusp of a revolution, an exhilarating fusion of human and artificial intelligence (AI) agents. This transformative wave promises to be nothing short of spectacular, yet it also poses a daunting challenge for those unprepared. I confess, the pace of change is staggering, particularly due to:

  • The synergistic dynamism between Artificial Intelligence and Intelligent Automation (AiA).
  • The rapid and widespread adoption of AiA by knowledge workers.
  • The significant influence of AiA reshaping our digital workforce.

In a recent discourse with Lex Fridman, Sam Altman peeled away layers from the complexities of workforce transformation, diving into the realm of AI agents. Despite the turbulence surrounding OpenAI, Altman highlighted that GPT-5 would pioneer the widespread creation of AI agents, setting off a domino effect in the industry. What doors will this open for the digital workforce?

Flash Forward:

Ai agents and Human in a digital workforce

Picture a future where AI agents are integral to your team, taking on tasks and assigning them – not just to other AIs but to humans as well. With GPT-5’s advent, these agents are poised to undertake certain tasks autonomously, marking a significant leap in AI’s capacity to operate independently of human oversight. It will begin with essential, straightforward tasks and gradually expand to a broader functional spectrum, encompassing various interconnected modalities such as text, speech, vision, and video.

And what value will this add to knowledge workers today? In marketing, AI agents will manage data analysis, report creation, email sorting, programming, and even orchestrating marketing strategies, bestowing the invaluable gift of time. But what actions must we take – or avoid – to leverage this gift effectively?

Stop Overfocusing on Tech Innovations:

While staying updated is crucial, avoid becoming consumed by technological advances unless you’re in product development. The race for the latest model or parameter count is a distraction for consumers of AiA.

Start Dissecting Your Workday:

Deconstruct your activities into distinct tasks. AI agents will substitute tasks, not positions, but ignorance of your own workflows could leave you vulnerable to displacement.

Maintain Digital Workforce Expansion:

For organizations, continue the push for digital integration. This past decade has seen immense digital transformation – it’s time to literally digitize human roles. Understanding and documenting workflows are imperative, as is investing in AiA literacy for your workforce.

Though these suggestions may seem straightforward, their implementation is what distinguishes pioneers from stragglers in this AI-driven era.

Digital Task-Based Workforce

A successful company will at start the day with valuable information.  After completing a bunch of tasks, that company ends the day with even more valuable information.  And from valuable information comes profit and growth.

Today, the digital workforce is growing to take on more connected and complex tasks. And with knowledge workers are applying artificial intelligence (AI) and intelligent automation (IA), or AiA, to complete tasks in their day to day, digital task-based workforce is growing at exponential rates.

Digital Task-Based Workforce

Task is the unit of work

So envision a workforce where the majority of tasks are digital. A massive ecosystem of small to medium sized connected tasks. Traditionally, knowledge workers spearheaded these tasks. It is their job to complete these tasks by leveraging their expertise across various domains. However, as AiA capabilities grow, this workload is being shared. The digital workforce is morphing into a landscape with a harmonious blend of human intellect and machine efficiency. This partnership has not only accelerated task completion by the digital workforce but also introduced innovative solutions to complex challenges.

Looking ahead to AI Agents

Within a task-based workforce, distinguishing between tasks completed by humans or machines becomes less relevant. But AiA enabled humans are the next first important step towards this reality. That requires people to get comfortable working with AI agents now. And the curious will lead the way.

As the digital workforce matures, humans will cede more control to AI agents. Trusting them with more and more tasks and shifting into a partnering relationships with smart capable AI agents with experience and expertise under their belt. And of course, in some cases the AI agents will lead with humans in support. We’re not there yet, but it’s only a matter of time, and only for jobs we’d consider simple fit for machines. But for now, test ahead, and start learning how AI agents can help you in your day to day.

Knowledge Workers in the AI Age

Knowledge workers use their intellectual skills to create, process, and share information. Scientists, architects, engineers, lawyers, teachers, analysts, planners, and software developers are just a few examples of knowledge workers that think for a living.  They are key contributors to innovation and problem-solving in various industries, leveraging their expertise to drive decisions and strategies.

However, the role and responsibilities of knowledge workers is evolving fast with the accelerated adoption of artificial intelligence (AI) and intelligent automation (IA) – AiA – into future digital workforce.

We’ve heard that 50% of knowledge workers have used AiA in some form or other, directly or indirectly. Moreover, that number increases to over 90% within marketing and advertising.  Interestingly, this usage is not driven by corporate policy.  It’s more self-taught and home schooled.  Nevertheless, AiA usage is on the rise.

Knowledge Workers in Marketing and Advertising

marketing knowledge worker

In marketing, knowledge workers craft messages that resonate with audiences for advertising outcomes. These creatives, from copywriters and designers to strategists and analysts, harness their expertise to develop campaigns that captivate and persuade. As the intelligence age advances, AiA is transforming knowledge worker roles and responsibilities, amplifying creativity and efficiency.

Let’s be clear here, AiA’s influence is profound across the end-2-end creative process.  We are automating routine tasks. Prioritizing the best use cases for specific AI tools from analyzing consumer data and predicting trends to generating intelligent outputs to inform targeted marketing strategies. This enables marketers to craft personalized campaigns that speak directly to individual preferences, significantly enhancing engagement and conversion rates.

Now an advertising agency might employ AI to generate multiple versions of an advertisement, each tailored to different audience segments based on online behavior and preferences. But that’s a lot of additional content.  However, the ambition is to streamline the creative process so that each ad variation is optimized for maximum impact.  It also shifts the certain responsibilities of knowledge workers from creator to curator in the creative process.  And all this what-if ad impact analysis is done at design time, before the ad is activated, if at all.  Yet again adding review and approve responsibilities to the knowledge worker. 

And AiA adoption is not a given

Adoption is critical for sustainable change.  HR departments are in the people trenches fielding the inevitable structural changes to digital workforce of the future. Here are a couple of major and minor organizational AI trends:

Major TrendMinor Trend
Investment increasingConfidence in right AiA investment bets
Likelihood of people replacementEmployee support for change
Workforce transformationExpertise for transformation

Basically, you hear a lot noise about the major trends and less about the minor trends. However, we need to pay more attention to the minors to bring favorable outcomes for the majors, increasing AiA adoption across the organization.

What is AiA

I’ve always been interested in getting things done (GTD). Both professionally and personally. Being productive with a focus on the execution side is important, but not without embracing the rise of intelligence in the workplace.

Right now, Intelligent Automation (IA) has my professional focus.  To drive better client outcomes by supporting knowledge workers you must bring that right blend of methodology and technology.  And you guessed it, AI is that technology making the difference today.  AI amplifies the impact of IA.  AiA, the synergy of AI and IA has intelligence at the center of everything.


The Rise of the Digital Workforce

The digital workforce is predominantly made up of human workers. However, AiA will add more and more automated processes, increasing the size of digital workforce.  But more importantly, the shape and structure of the digital workforce will change significantly and permanently.  And the role of humans within the digital workforce will be even more critical.  We have officially entered the future of work discovery phase.  No one knows what that digitial workforce mix will be.  We are all learning together.

However, ignorance is no excuse.  What we do know is that intelligent knowledge workers, those that can work with AiA, will have more opportunities within the shifting and growing digital workforce, than those that can’t. 

AiA gifts knowledge workers with time to perform more human-centric tasks.  By delegating the manual and mundane to machines, we can embrace the creative and innovative to drive intelligent outcomes for others. 

So, where are we today?

Every business is different.  Culture, strategy, talent, operations, innovation, data, technology, just to name a few, and their inter-dependencies, define an organization.

However, the race is to digitize everything, everywhere, all at once, and intelligently.   Do you know what you do?  Are your processes undocumented, documented, automated, and/or integrated?  What’s your mix? Is your scope internal, external, with or without partners? And your data, who owns it?  How do you share, secure, syndicate, sanitize, and simplify it?  Do you know how to monetize what you have?  Have you separated what you sell from effort required to deliver it?

There are just a few questions to address before you can scale your business with AiA.

And the next step is…

Shared awareness.  We are in the intelligence revolution. Steam, mass production, and digital are industrial revolutions that have been and gone.  Now the transformation towards intelligent connected systems, fueled by data, powered by machines, and directed by humans is in full swing. Our forward-thinking leaders are placing big bets in AiA with the clear understanding that the intelligence landscape is critical to future growth and survival.

So, empowered execution follows shared awareness.  Those at the tip of shift have started executing, but not underestimate the amount of effort it requires to reach shared awareness.  And that’s where most of us are playing now.

Help writing help documentation

Creating help documentation is a skill. It is usually not one that is done well by those responsible for building the solution. But you can get help writing help documentation. I recently discovered Synthesia and Scribe, two AI solutions that make the creation of help docs easier. 

Enter Synthesia 


Who doesn’t like video? Synthesia makes it super simple to create help documentation using its library of avatars. Just pick an avatar, enter text in the language of your choosing, and have it speak to your solution. Then generate, share the video, and your done! 

Enter Scribe 

I love Scribe. Just hit the generate button and it will capture how you move about the screen. It’s smart, automatically creating how-to guides. Saving you the hassle of screenshots, annotating them, and then assembling them. Scribe shortens the distance between creation to curation for documentation.  

About Cleve Gibbon

Hey, I’m Cleve and I love technology. A former academic that moved into fintech to build trading platforms for investment banks. 20 years ago I switched to marketing and advertising. I joined a content technology spin-off from the Publicis network that was bought by WPP in 2014. I'm now at Omnicom. These pages chronicle a few of things I've learnt along the way…

My out-of-date cv tells you my past, linked in shares my professional network and on twitter you can find out what I'm currently up to.