Session brief
Where AI is superhuman: The right jobs to automate with LLMs
Overview
This talk explores how Large Language Models (LLMs) can automate tasks in the workplace, focusing on identifying jobs where AI can be superhuman. It posits that LLMs excel at data transformation, synthesis, and reasoning, making them particularly disruptive for high-volume, low-complexity tasks. The presentation suggests that AI automation will shift organizational structures from pyramid shapes to inverted pyramids or diamonds, with fewer entry-level roles and more advanced or managerial positions.
Who should watch
- AI engineers and builders exploring automation opportunities.
- Product Managers evaluating where to focus AI development.
- Executives seeking to understand AI's impact on business functions and organizational structure.
- Founders deciding on startup niches within AI workflow automation.
- Anyone interested in the practical application of LLMs beyond basic co-pilot functions.
Key takeaways
- LLMs are powerful for transforming data, synthesizing information, and approximating reasoning, solving historical blockers in workflow automation.
- Jobs can be categorized on a spectrum of volume and complexity, with high-volume, low-complexity tasks being the most susceptible to complete automation by LLMs.
- In these high-volume, low-complexity roles, LLM systems can outperform humans significantly, not just in accuracy but in sheer capacity.
- Examples like security operations (Drop Zone AI) and customer engagement (Amp) demonstrate agentic systems handling end-to-end workflows, outperforming rule-based systems and providing personalized experiences.
- AI automation will likely lead to organizational restructuring, with a decrease in junior roles and an increase in roles focused on reviewing AI outputs, maintaining systems, and higher-level strategic work.
- The competitive landscape for AI automation is shifting from human versus AI to AI versus previous generations of rules-based software.
- LLMs can uncover new customer cohorts and insights, as seen with a food delivery company identifying late-night snackers through personalized messaging experiments.
- Understanding the task breakdown within jobs and the spectrum of complexity and volume is crucial for identifying effective AI automation targets.
Notable quotes
*LLMs effectively solve this problem so they're really powerful.*
*High volume relatively low complexity jobs will be the most transformed by LLMs because that's where they're already superhuman.*
*Your competition is no longer AI versus human it's AI versus previous generation of rules-based software.*
Unofficial community note. Prefer the recording for nuance.