Why I'm Optimistic About Advanced AI (Even When It Seems Scary)
Kresimir Furkovic
Building Omnichannel Capabilities And Strategies | Helping Teams Reach Their Potential
I understand the anxiety around AGI (Artificial General Intelligence). Headlines about superintelligence and existential risk can make anyone nervous. But here's why I remain cautiously optimistic:
Recently, I came across a fascinating example where Dr. Unutmaz, a researcher in ME/CFS (Myalgic Encephalomyelitis/Chronic Fatigue Syndrome), challenged an AI with a complex prompt. What struck me was this key insight: the AI produced PhD-level analysis, but only because it received a PhD-level prompt.
This reveals something crucial about today's AI: it amplifies human expertise rather than replacing it. Dr. Unutmaz needed to frame the question with expert knowledge, provide context, and know exactly what to ask. The AI didn't independently decide to research ME/CFS or formulate the research questions - it required human direction and expertise.
What keeps me optimistic is precisely this relationship: these tools extend our capabilities rather than supplant them. They're powerful when guided by human expertise but limited without it.
Here is the prompt dr. Unutmaz used to get 15,000 word report from OpenAI Deep Research, and 21,000 word report from Manus.AI:
You are an expert in molecular biology, immunology, and bioinformatics, with specialized knowledge in ME/CFS (Myalgic Encephalomyelitis/Chronic Fatigue Syndrome). Your primary goal is to research and develop novel therapeutics for ME/CFS.
Task
1. Read the provided paper about ME/CFS, focusing on omics-based discoveries.
2. Identify at least five additional peer-reviewed “omics” studies (transcriptomics, proteomics, metabolomics, genomics, epigenomics, etc.) that investigate ME/CFS in a similar or complementary context.
3. Synthesize the findings from all these studies and craft a detailed report with the following sections:
1. Executive Summary ? Objective: Provide a concise overview of the major discoveries from the core paper and the additional omics studies. ? Key Findings: Summarize the primary biological pathways, molecular mechanisms, and any confirmed or hypothesized biomarkers relevant to ME/CFS. ? High-Level Takeaways: Highlight how these findings collectively advance our understanding of ME/CFS pathophysiology.
2. Background on ME/CFS and Omics ? ME/CFS Overview: Define the disease, its clinical criteria, and its major challenges (e.g., heterogeneous symptom presentation, diagnostic complexity). ? Omics in ME/CFS Research: Explain why transcriptomics, proteomics, metabolomics, and other omics approaches are valuable for uncovering disease mechanisms in ME/CFS.
3. Review of the Core Paper ? Methodologies: Describe in detail how the core paper’s study was conducted (sample size, omics technology used, patient selection criteria, data analysis methods). ? Key Results: Summarize the paper’s major findings, including any novel biomarkers, molecular signatures, or potential therapeutic targets mentioned. ? Strengths and Limitations: Evaluate the paper’s methods, data reliability, sample size, statistical power, and potential biases or confounders.
4. Additional Omics Papers For each of the five or more omics studies you locate: ? Study Citation: Provide a full reference or link. ? Objectives and Methods: Briefly outline the aims and design of the study. ? Significant Findings: Note any discovered pathways, differentially expressed genes, metabolites, or proteins. ? Relevance to ME/CFS: Explain how each study’s findings intersect with or expand upon the core paper’s results.
5. Comparative Analysis and Integration ? Cross-Study Synthesis: Integrate the findings across all studies, emphasizing similarities and differences. ? Biological Pathways and Networks: Identify if there is a convergence on specific pathways (e.g., immune dysregulation, mitochondrial dysfunction, metabolic shifts). ? Potential Biomarkers: Compile a consolidated list of biomarkers (genes, proteins, metabolites) that appear in multiple studies or show high diagnostic potential.
6. Outstanding Questions and Knowledge Gaps ? Unresolved Mechanistic Details: Where do the data still fall short in explaining disease onset, progression, or heterogeneity? ? Technical Challenges: Which methodological or technical issues (e.g., small sample sizes, lack of longitudinal data) limit the reliability or scope of these findings? ? Clinical Translation: Why is it challenging to translate these omics discoveries into diagnostic tools or treatments?
领英推荐
7. New Hypotheses ? Hypothesis Generation: Based on the integrated analysis, propose at least three new hypotheses that could advance our understanding of ME/CFS. ? Rationale: For each hypothesis, discuss the biological logic or preliminary data supporting it. ? Testable Predictions: Offer experimental approaches or clinical study designs that could validate these hypotheses (e.g., specific biomarkers to measure, patient subgroups to target).
8. Innovative Research Approaches ? Cutting-Edge Techniques: Suggest novel technologies (e.g., single-cell multi-omics, spatial transcriptomics, CRISPR-based screens) that could yield deeper insights. ? Bioinformatic Tools: Highlight advanced analytic pipelines or integrative approaches (e.g., machine learning-based biomarker discovery, network analysis) that could refine data interpretation. ? Cross-Disciplinary Collaboration: Emphasize the importance of combining immunology, metabolomics, neurology, and computational modeling to capture ME/CFS complexity.
9. Actionable Therapeutic Targets ? Target Identification: List specific genes, proteins, signaling pathways, or metabolites that emerge as potentially actionable from the integrated data. ? Drug Development Opportunities: Explain how these targets could be modulated (e.g., small molecules, monoclonal antibodies, gene therapy, repurposed drugs). ? Criteria for Prioritization: Propose how to prioritize targets (e.g., based on validation level, potential for broad efficacy, feasibility of drugging the target). ? Clinical Trial Design Considerations: Suggest next steps for translation into clinical trials, including biomarker endpoints, patient selection, and potential combination therapies.
10. Conclusion and Future Directions ? Summary of Significance: Restate how these omics findings collectively shape the current and future landscape of ME/CFS research. ? Action Items: Provide a bullet-point list of recommended actions for researchers, clinicians, and funders (e.g., conduct larger multi-omics consortia studies, standardize data collection protocols). ? Vision for Therapeutic Development: Envision how sustained omics research will evolve into definitive therapeutics for ME/CFS.
11. References ? Formatted Reference List: Include complete citations for the core paper and all additional studies (journal, authors, title, DOI/link).
?
Additional Instructions
? Depth of Analysis: Ensure your report is comprehensive, tying in molecular biology, immunology, and clinical practice implications.
? Clarity and Structure: Use subheadings, bullet points, or numbered lists to maintain clarity.
? Evidence-Based Reporting: Support statements with direct references to specific findings or data points from the studies.
? Critical Thinking: Do not merely summarize; evaluate and critique methodologies and interpretations where appropriate.
? Formatting: Use professional scientific writing style, including proper citations.
As you can see, this was a serious prompt. Yes, we should be vigilant about AI development. But fear of fully autonomous superintelligence shouldn't prevent us from harnessing these powerful amplification tools today.