The AI Expectation Gap: Why We're Stuck at Level 1, Dreaming of Level 3

If you are following AI developments, we're definitely living in exciting times. AI seems to be rapidly evolving, with promises of revolutionising industries and transforming the nature of our work lives. Yet, amidst all the excitement and hype, a stark reality emerges: a significant gap exists between our expectations of AI and its current capabilities.

Take a look at the chart from Artificial Workflow:

We're currently at?Level 1: Independent Agents. AI can perform single tasks, augmenting human capabilities in limited ways. Think chatbots answering basic customer queries or language models like ChatGPT helping write social media posts.

There are extensions into Level 2: Action Agents characterised by automated task execution through pre-programmed workflows by leveraging platforms like Zapier or AutoGen Studio. These are valuable capabilities solving real-world business problems, but they lack the independent reasoning and problem-solving associated with higher level AI advancements.

But the hype cycle has kicked in. We're already hearing ‘people’ talk about the possibilities of?Level 3: Autonomous Agents?- AI systems capable of independent operation, managing complex tasks and even entire projects with minimal human intervention. Reality check - we are a long way off this being available.

So, why the disconnect?

Several factors contribute to this expectation gap:

• The "Shiny Object" Syndrome:?We're captivated by the allure of sophisticated AI, often overlooking the foundational work needed to get there.
• Marketing Hype:?Exaggerated claims and promises of "AI-powered everything" fuel unrealistic expectations, masking the complexities of true AI development.
• Limited Understanding of AI's True Potential:?We haven't fully grasped the nuances of different AI stages and their organisational implications, leading to misplaced expectations.

Navigating the Levels of AI Evolution

OpenAI has provided a framework to gauge the progression towards artificial general intelligence (AGI), and identify current AI capabilities at Level 1. Here is a description of the levels and their potential impact:

1. Conversational AI: The current stage, where AI can interact in conversational language, exemplified by tools like ChatGPT. It is widely used for customer support, content creation, and more, with significant user adoption since its launch.
2. Reasoning AI: The next anticipated stage involves systems that can perform basic problem-solving tasks at a level comparable to a human with a doctorate, but without external tools. This stage represents a significant leap in AI capabilities.
3. Autonomous AI: At this level, AI systems, or "agents," can operate autonomously on behalf of users for extended periods. These agents would require minimal human intervention, functioning like reliable team members or suppliers.
4. Innovating AI: These "innovators" can independently develop new innovations, improving processes and thinking critically about achieving goals more effectively. This stage would significantly enhance business operations and strategic planning.
5. Organisational AI: The final stage involves AI systems capable of performing the work of an entire organisation, coordinating various functions, making improvements, and running operations without human involvement. This represents the ultimate goal of AGI.

Bridging the Gap: Knowledge is Power

The key to navigating this AI landscape lies in understanding where we are and where we're headed. Here's some suggestions to bridge the expectation gap:

• Educate Yourself:?Invest in understanding different AI levels and their potential impact on your industry and organisation.
• Manage Expectations:?Engage in realistic conversations about AI capabilities and timelines, tempering hype with grounded assessments. Fact check someone hyping AI ?by asking what level do they think we are currently at?
• Strategic Planning:?Align your AI strategy with realistic goals, starting with achievable Level 1 and Level 2 implementations.
• Focus on Human-AI Collaboration:?Develop a workforce adept at working alongside AI, fostering a culture of collaboration and continuous learning aligned with Level 1 and Level 2.

Conclusion

Understanding where we are (still firmly in Level 1) and what is possible is crucial for setting realistic expectations and strategies. By acknowledging the current capabilities and incremental nature of AI advancements, organisations can better manage expectations and strategically plan for the future.

So, the next time you hear grand promises about AI's potential, remember: we are just at the beginning. Recognise the hype, understand the reality, and prepare your organisation for the gradual yet profound evolution of AI.