23. Agentic AI ROI From Automation to Decisions

23. Agentic AI ROI: From Automation to Decisions 🔗 **Source:** [23. Agentic AI ROI: From Automation to Decisions](https://www.youtube.com/watch?v=PH6KNZykjX4) > [!note]+ Document Notes > tags: [agentic-ai, digital-labor, ai-roi] > author: Scott Hebner > source: theCUBE Research > reference: https://www.youtube.com/watch?v=PH6KNZykjX4 > date created: 2024-12-19 14:30 > > # 🏷️ Title: Agentic AI ROI: From Automation to Decisions >  ## 🧰 Action Inbox - [H1] Deploy AI agents in parallel with existing workforce for gradual trust building — 18:23 #trust #deployment - [H2] Focus on integration capabilities not just innovation when scaling agents — 25:47 #integration #scaling - [H3] Start with quick wins and reliable use cases before expanding agent capabilities — 22:15 #strategy #rollout ## 🧭 Table of Contents - [📝 Summary](#-summary) - [📖 Main Content](#-main-content) - [🗒️ Notes](#️-notes) - [📚 References](#-references) - [🗣️ Paraphrase](#️-paraphrase) - [📰 Newsletter](#-newsletter) - [✅ Action Items](#-action-items) - [🧪 Hypotheses](#-hypotheses) - [📊 Methodology](#-methodology) - [❓ Questions](#-questions) - [🔮 Future Research](#-future-research) - [💡 Implications](#-implications) - [🧠 Learning Styles Assessment](#-learning-styles-assessment) - [📊 Accessibility Assessment](#-accessibility-assessment) - [🌍 DEIB Assessment](#-deib-assessment) - [✅ Logical Validity](#-logical-validity) - [🛡️ DARVO Analysis](#-darvo-analysis) - [🎯 Conclusion](#-conclusion) ### 🔗 References - [Next Frontiers of AI Podcast](https://www.youtube.com/watch?v=PH6KNZykjX4) - [DoozerAI Website](https://doozer.ai) ## 📝 Summary - The podcast explores findings from the Agentic AI Futures Index survey showing 61% of organizations plan to deploy AI agents within 18 months - Investment is shifting from generative AI to agentic AI capabilities that enable reasoning and decision-making rather than just content creation - Trust emerges as the critical barrier to ROI, with only 49% expressing high trust in AI agents for decision-making - Paul Chada from DoozerAI shares real-world lessons on deploying digital workers alongside human teams - Key recommendation is to start small with reliable use cases and build trust gradually through parallel deployment ## 📖 Main Content The enterprise AI landscape is experiencing a fundamental shift from generative AI experimentation to agentic AI deployment. According to new survey data from theCUBE Research's Agentic AI Futures Index, 61% of organizations plan to deploy AI agents within the next 18 months, marking a transition from demonstrating potential to delivering tangible ROI. This represents a move from a software automation market to a digital labor market, where AI agents function as digital coworkers rather than mere productivity tools. Labor expenses typically represent 60% of enterprise costs, making the potential impact of digital labor transformation significantly larger than previous automation waves. However, trust emerges as the primary barrier to success. Only 49% of AI professionals express high confidence in AI agents for decision-making tasks, creating a gap between investment intentions and deployment readiness. Organizations are investing heavily in AI reasoning capabilities (73% planning significant investments) but struggling with the cultural and technical challenges of building reliable, explainable systems. ## 🗒️ Notes #### ⏱️ [00:00](https://www.youtube.com/watch?v=PH6KNZykjX4?t=0) — [H1] Introduction to Agentic AI Shift - Key idea: Moving from generative AI content creation to agentic AI decision-making - Evidence: "The real story now is the shift to agentic AI, digital co-workers that go beyond creating content and automating tasks to actually reasoning, making decisions, and collaborating in the flow of work" #### ⏱️ [02:30](https://www.youtube.com/watch?v=PH6KNZykjX4?t=150) — [H2] Paul Chada's Background and DoozerAI Origins - Key idea: Founded DoozerAI after recognizing LLMs could reach previously untouchable ROI use cases - Evidence: "We realized that it could actually reach some of these hard-to-reach ROI use cases that previously had been sort of untouchable" #### ⏱️ [05:45](https://www.youtube.com/watch?v=PH6KNZykjX4?t=345) — [H2] Digital Labor Transformation Inevitability - Key idea: 90%+ believe today's leaders will be last to manage human-only workforce - Evidence: "Today's leaders will be the last ones to manage a human-only workforce" #### ⏱️ [08:12](https://www.youtube.com/watch?v=PH6KNZykjX4?t=492) — [H2] Parallel Deployment Strategy - Key idea: Run agents alongside human workers to build trust gradually - Evidence: "Build an agent, run it in parallel with your existing workforce, and then you can sort of ease up on the human oversight over time" #### ⏱️ [10:30](https://www.youtube.com/watch?v=PH6KNZykjX4?t=630) — [H1] Survey Data: 61% Planning Agent Deployment - Key idea: Significant portion of enterprises moving beyond pilots to deployment - Evidence: "61% said that they plan to deploy AI agents over the next 18 months" #### ⏱️ [12:15](https://www.youtube.com/watch?v=PH6KNZykjX4?t=735) — [H2] Measured Approach to Agent Deployment - Key idea: Successful organizations taking cautious, infrastructure-first approach - Evidence: "They're choosing some of the easy wins to get the technology in place" #### ⏱️ [14:45](https://www.youtube.com/watch?v=PH6KNZykjX4?t=885) — [H2] Platform vs Build-Your-Own Success Rates - Key idea: MIT study shows platform approach more successful than building internally - Evidence: "The ones that tried to build it themselves were more likely to fail than the ones that just bought a platform" #### ⏱️ [16:20](https://www.youtube.com/watch?v=PH6KNZykjX4?t=980) — [H1] Investment Focus on AI Reasoning - Key idea: 73% planning significant investment in AI reasoning and decision intelligence - Evidence: "73% of enterprises are planning either significant or strategic investments in enabling AI agents that can reason" #### ⏱️ [18:30](https://www.youtube.com/watch?v=PH6KNZykjX4?t=1110) — [H2] Knowledge Worker Focus - Key idea: Organizations targeting high-value, expensive knowledge worker roles - Evidence: "The money's probably flowing mostly into less chat bots, more into knowledge worker tasks" #### ⏱️ [20:15](https://www.youtube.com/watch?v=PH6KNZykjX4?t=1215) — [H1] Decision-Making as Primary Agent Activity - Key idea: 62% want agents to assist in making better decisions - Evidence: "62% actually want these agents to help them make better decisions" #### ⏱️ [21:45](https://www.youtube.com/watch?v=PH6KNZykjX4?t=1305) — [H2] Financial Reporting Agent Success Story - Key idea: Five-month deployment demonstrates reliability through parallel operation - Evidence: "That one's been running for about five-and-a-half months now... they're saying, 'No, it's really fine'" #### ⏱️ [23:10](https://www.youtube.com/watch?v=PH6KNZykjX4?t=1390) — [H1] Trust as Currency of Innovation - Key idea: Only 49% have high trust in agents for decision-making - Evidence: "Only 49% have a high degree of trust in these agents to make decisions" #### ⏱️ [24:30](https://www.youtube.com/watch?v=PH6KNZykjX4?t=1470) — [H2] LLM Malleability Problem - Key idea: ChatGPT's tendency to change answers undermines trust in agent decisions - Evidence: "You could ask it something, and it gives you an answer. And then you say, 'But I was thinking it was this,' and it goes, 'Oh, you're right. It is this'" #### ⏱️ [26:00](https://www.youtube.com/watch?v=PH6KNZykjX4?t=1560) — [H3] Start Small, Build Trust Strategy - Key idea: Gradual rollout with proven successes builds organizational confidence - Evidence: "Start small, build upon the successes, gain adoption, and then that trust will increase" ## 📚 References - [DoozerAI Platform](https://doozer.ai) - [MIT Study on AI Agent Implementation](URL not provided) - [theCUBE Research Agentic AI Futures Index](URL not provided) - [Digital Labor Transformation Index](URL not provided) ## 🗣️ Paraphrase - Enterprise AI is shifting from content generation tools to decision-making digital coworkers that work alongside human teams - Trust remains the biggest barrier despite heavy investment, requiring gradual deployment strategies that prove reliability over time - Organizations are moving from software automation mindset to digital labor hiring, focusing on knowledge worker augmentation rather than simple task automation ## 📰 Newsletter **Agentic AI Takes Center Stage: From Hype to ROI-Driven Reality** The enterprise AI landscape is experiencing a seismic shift. New research from theCUBE's Agentic AI Futures Index reveals that 61% of organizations plan to deploy AI agents within 18 months—not just for automation, but as genuine digital coworkers capable of reasoning and decision-making. **Key Insight:** We're transitioning from a software market to a labor market. When 60% of enterprise costs are labor-related, the ROI potential of digital workers far exceeds previous automation waves. **The Trust Challenge:** Only 49% of AI professionals highly trust agents for decision-making. Success requires starting small, running agents in parallel with human workers, and building confidence through demonstrated reliability. **Practical Next Steps:** Focus on integration over innovation, choose platform solutions over building internally, and prioritize explainable AI that teams can understand and trust. Resources: [Learn WordPress](https://learn.wordpress.org/), [Developer Docs](https://developer.wordpress.org/), [WP-CLI](https://wp-cli.org/) ## ✅ Action Items #### DevRel - Developer advocacy: Create agent development workshops and demos showcasing ROI metrics - Education: Develop curriculum on agentic AI patterns and trust-building strategies - Developer experience: Build SDKs that emphasize explainability and observability - Developer marketing: Campaign highlighting real-world agent success stories - Community management: Facilitate discussions on agent deployment best practices #### Community - Maintainers: Develop governance templates for AI agent integration - Contributors: Create first-issue programs for agentic AI contributions - Sponsors: Transparent reporting on AI investment ROI and impact - Ecosystem: Cross-project coordination on agent interoperability standards ## 🧪 Hypotheses - Organizations deploying agents in parallel with human workers will achieve higher trust scores than those attempting full automation - Platform-based agent solutions will demonstrate 2x higher success rates than custom-built implementations - Explainability features will correlate strongly with user adoption rates of AI agents - Trust scores will increase significantly over 6-month measurement periods with proper deployment strategies ### 📊 Methodology - Longitudinal study comparing parallel vs full automation deployment outcomes - Platform vs custom solution success rate analysis across 100+ implementations - Trust measurement surveys conducted quarterly with agent users - Explainability feature usage correlation with adoption metrics ### ❓ Questions - How will agent trust scores evolve as deployment strategies mature over the next 18 months? - What specific explainability features most effectively build user confidence in agent decisions? - How do different industries vary in their readiness to adopt agentic AI for knowledge work? ### 🔮 Future Research - Development of standardized agent trust measurement frameworks - Investigation of industry-specific agent deployment patterns and success factors - Analysis of agent-human collaboration models and their impact on productivity metrics ### 💡 Implications - Practice: Shift from software procurement to digital hiring mindset required - Policy: New governance frameworks needed for agent decision-making accountability - Security: Agent behavior monitoring and anomaly detection becomes critical - Education: Workforce training on human-agent collaboration patterns essential - Funding: Budget allocation shifting from software licenses to digital labor capacity ### 🧠 Learning Styles Assessment Reading level estimate: Flesch-Kincaid Grade Level 12 #### 👀 Visual learners - Present: Survey data charts showing investment priorities and trust levels - Improve: Agent workflow diagrams, decision tree visualizations, trust-building timelines #### 🔊 Aural learners - Present: Full podcast audio with expert discussion and real-world examples - Improve: 3-minute executive summary audio, key quote highlights #### 🗣️ Verbal learners - Present: Clear definitions of agentic AI, extensive examples and analogies - Improve: Glossary of technical terms, comparison frameworks #### 🧩 Kinesthetic learners - Present: Step-by-step deployment strategies, actionable checklists - Improve: Interactive agent deployment calculator, hands-on workshops #### 🧠 Logical learners - Present: Statistical data, survey methodology, causal relationships - Improve: Decision matrices for agent selection, ROI calculation frameworks #### 🧑‍🤝‍🧑 Social learners - Present: Expert interview format, community discussion points - Improve: Peer group deployment case studies, collaborative planning templates #### 🧘‍♀️ Solitary learners - Present: Comprehensive written analysis, reflection questions - Improve: Self-assessment tools for agent readiness, individual learning paths #### 🌿 Naturalistic learners - Present: Ecosystem thinking about agent-human collaboration - Improve: Organizational change lifecycle models, adaptation strategies ### 📊 Accessibility Assessment #### Screen reader semantics - Present: Clear heading structure, descriptive chart alt text - Improve: ARIA labels for survey data visualizations, table summaries #### Keyboard navigation - Present: Standard video controls, logical tab order - Improve: Skip links for lengthy transcript, keyboard shortcuts #### Contrast and readability - Present: High contrast text and charts - Improve: Verify color-blind accessibility for data visualizations #### Text resizing and reflow - Present: Responsive design supporting zoom - Improve: Test reflow at 200% zoom, ensure no content clipping #### Media and captions - Present: Full transcript available - Improve: Add timestamps to transcript, downloadable SRT file #### Forms and interactions - Present: Accessible video player controls - Improve: Keyboard-accessible chart interactions ### 🌍 DEIB Assessment #### Cultural breadth - Present: California and Ireland perspectives represented - Improve: Global enterprise examples, multi-regional case studies #### Inclusive language - Present: Gender-neutral technical terms, defined jargon - Improve: Plain-language summaries of complex concepts #### Neurodiversity support - Present: Structured content with clear headings and timestamps - Improve: Visual break indicators, consistent formatting patterns #### Representation - Present: Diverse professional backgrounds and experiences - Improve: Include perspectives from various organizational levels #### Psychological safety - Present: Respectful discussion of challenges and failures - Improve: Emphasize learning from mistakes, supportive tone ### ✅ Logical Validity #### Strong reasoning - Deductive strength: Survey data supports conclusions about enterprise readiness - Inductive strength: Case studies generalize to broader deployment strategies - Causal reasoning: Clear links between trust and ROI outcomes - Analogical strength: Browser wars comparison effectively illustrates LLM role #### Fallacies - Hasty generalization: Limited case studies presented as universal patterns - provide more diverse examples - Appeal to authority: Heavy reliance on survey data without methodological details - include sample size and methodology #### Tone of voice - Overall tone: Professional, data-driven, and cautiously optimistic with realistic acknowledgment of challenges ### 🛡️ DARVO Analysis - Deny: No denial patterns observed - Attack: No attack patterns on critics or skeptics -