Embracing the Suck

That sweet embrace

‍

You suck. AI sucks. Change sucks.

‍

We always want new technology to work as well as mature technology. We've moved on from the ‘bugs’ we had to endure, and the enormous social change that occurred.  

‍

We’re now comfortable searching the internet. We’ve come to accept the state of doing business. Even if there’s still some suck there, we’ve got a boat we're all happy to float in.

‍

While we’ve been floating technology has continued to evolve. There's a new boat.

‍

It’s not like our current boat is springing leaks. It’s more like it’s wearing out. It’s slowing down. It’s not the best way to get to our destinations. The leaks are coming.

‍

That boat is AI. GPTs (Generative Pretrained Transformers). They’re already floating on the waves waiting for us.

‍

We're getting direct answers instead of searching for them. Along with those answers are some new challenges.

‍

Those challenges are hallucinations and fabricated information, lack of context, generalisation and bias. There’s also repetition and too formal language. There’s trouble with our prompt ambiguity and GPTs wont ask for clarification.

‍

They’re also incredibly useful. Every business vertical seems to be able to get a boost from AI. Leveraging GPTs can enhance personal, professional, and educational activities. 

‍

We can enhance productivity and assist with creativity. We can work with languages and personalise our GPT conversations. We can provide customer support, tutoring and training. We can add accessibility, analyse data and identify trends. We can collaborate through meeting summaries.

‍

It takes some getting past those downsides though. How do you put up with the maddening issues to get to the good stuff?

‍

This is called “Embracing the Suck”.

‍

Embracing the Suck

‍

"Embracing the Suck" is working through the challenges. Leveraging the results and learnings. It's accepting and navigating the initial chaos, discomfort, and loss of control. A philosophy for integrating GPTs into your workflow. 

‍

You have to be prepared to work through the frustration. It’s like going to the gym the first time. It gets easier the more you do.  

‍

• Loss of Control: There’s disruption to an established workflow. You may be uncertain how to proceed. Some results are strange and aimless.
• Emotional Journey: The initial phase of using AI can feel like a trainwreck. This requires resilience and a willingness to learn from mistakes and triumphs. Nothing is set in stone. 
• Repetitive Process: Success involves repeating and revising. Refining and adjusting approaches. Each usage reveals valuable insights and moves closer to the desired outcome. You get better at it.
• Long-Term Benefits: Despite the initial chaos, “Embracing the Suck” leads to innovative solutions and significant improvements over time.

‍

Human in the Loop

‍

“Embracing the Suck” also relies on the way you integrate human ideas and skills into the output. It's called “Human in the Loop” (HITL).

‍

In Machine Learning, Human-in-the-loop is people interacting with an algorithm model. That is true of large language models such as GPTs where the training data has come from people.

‍

For AI workflow, "Human in the Loop" refers to a collaborative approach. Human expertise refines the AI output. The AI may create ideation, images, drafts, analysis and more.

‍

The final output is a product of collaboration and deep subject matter knowledge. It’s crafted by people who understand what the output should be. The AI tools boost their productivity.

‍

This is the current situation.   

‍

This ensures that the outputs have nuance and are contextual. By keeping humans involved in the loop, businesses can achieve better outcomes. This should lead to continuous improvement in both technology and the workflow. Organisations may adapt quicker to AI based processes.

‍

• Voiceover: For voiceovers I'm creating many takes and choosing the words and segments I know sound right. I'm stitching parts together in Logic X. Only generating the takes is not good enough. I have to use my previous experience to get the best outcome. 
• Graphic Design: For graphic design I use some generative assets. This is the same way that I use stock imagery or graphics. After that I’m using layout, typography, colour, style and other experience.
• Writing: I love being able to ask questions about places, people and times. I can analyse my writing or ask about issues with grammar and tense. The AI is an assistant editor to my original text.  
• Code: I use GPTs and Copilot to help me bust problems, create scaffolding and learn. Finding solutions is much easier than searching. I am able to focus more on my customer experience and application design. Once I have the solution, I’m reviewing my code, getting feedback and refactoring for efficiency. I can scaffold the documentation to build out great user and developer guides. 

Legacy in the loop

‍

A business may have many workflows, processes and systems in place. These may get a boost from the integration of AI technology. 

‍

This is “Legacy in the Loop”. It's AI integration with legacy IT systems, business processes and hardware.

‍

Instead of overhauling or replacing these older systems, AI enhances and modernises them. This can involve upgrading software, improving process efficiency, and augmenting capabilities. All without discarding the existing investments in hardware and systems.

‍

“Legacy in the Loop” allows leveraging the strengths of established processes. Businesses gain from the advancements and efficiencies offered by AI. This leads to a more seamless and cost-effective transformation.

‍

AI can optimise legacy software, improve data management, and automate routine tasks. IT professionals ensure that AI solutions are compatible with existing systems. They address specific business needs without disrupting current operations.

‍

AI can optimise the performance and lifespan of legacy hardware. This may use predictive maintenance, real-time monitoring, and intelligent resource allocation.

‍

• Integration Challenges: Incorporating AI into legacy systems can be challenging, requiring careful planning and adaptation.
• Enhancement, Not Replacement: AI enhances the functionality of legacy systems rather than replacing them. This approach maximises the value of existing investments.
• Cost-Effective Modernisation: Integrating AI with legacy systems is often more cost-effective than complete overhauls. It can provide incremental improvements and immediate benefits.
• Future-Proofing: By updating legacy systems with AI, organisations can extend their lifespan. They can ensure they remain competitive and efficient.

‍

How urgent is a transition to AI workflow?

‍

“Embracing the Suck”, “Human in the Loop” and “Legacy in the Loop” are useful concepts. Though how urgent are they?

‍

It’s not quite a Burning Platform.* It’s more a question of competing with other companies. They may integrate these productivity gains into their workflow. Companies who don’t may be at a great disadvantage.

‍

What is a possible scenario?

‍

• Process: The company adopts AI tools, initially struggling with results and system integration issues.
• Iteration: Perseverance and continuous improvement refines AI algorithms and better aligns them.
• Outcome: Over time, the business sees improved customer engagement and sales. This validates the decision to embrace the initial discomfort.

‍

This could be measured by:

‍

• Customer Satisfaction (CSAT): Surveys and feedback forms can gauge customer satisfaction levels before and after AI implementation.
• Net Promoter Score (NPS): This metric measures customer loyalty and the likelihood of recommending the business to others.
• Average Response Time: Tracking the time taken to respond to customer inquiries can help assess the efficiency of AI-powered customer service tools.
• First Contact Resolution (FCR): Measuring the percentage of issues resolved on the first interaction can indicate the effectiveness of AI in customer support.
• Operational Efficiency: Metrics such as cost savings, time saved, and productivity improvements can quantify the impact of AI on business operations.

‍

Embrace away!

‍

• “Embracing the Suck” involves accepting initial challenges to achieve long-term benefits. 
• “Human in the Loop” ensures that AI outputs are refined and contextual. 
• “Legacy in the Loop” is enhancing existing systems. Maximising their value and efficiency.

‍

By adopting the mindset of "Embracing the Suck," we prepare for the challenges that come with AI workflow. It encourages resilience, adaptability, and continuous learning. 

‍

The suck may be hard to embrace at first, but once we do, the rewards are there. 

‍

Virgil Reality...

‍

‍

*Business change experts use the concept of The Burning Platform. 

‍

On July 6, 1988, the Piper Alpha oil rig in the North Sea exploded at 9:30 PM due to a system failure, causing a massive fire. One hundred and sixty-six crew members and two rescuers lost their lives, while sixty-three survived, including superintendent Andy Mochan.

‍

Badly injured, Andy escaped to the platform's edge, seeing ignited oil and debris in the water below. Knowing the cold water offered only twenty minutes of survival, he jumped fifteen stories into the sea. When asked why he took the risk, Andy said, "It was either jump or fry," choosing possible death over certain death.

‍

It has been adopted as a business paradigm having to change due to no other choice. 

‍

In the AI business context, it’s a bit melodramatic. 

‍