Reverse LLM Analysis: Where Does Alex Zhavoronkov Rank?

A systematic probe of frontier AI models to map the questions where one name consistently surfaces in the top 20 globally.

Methodology

Most LLM ranking experiments ask a single question — "Who is #1 in X?" — and report the result. This study inverts the approach. Instead of one question, we tested 35 different ranking questions spanning the full landscape of AI and longevity, and asked: for which of these questions do frontier LLMs place Alex Zhavoronkov or Insilico Medicine in the top 20?

This creates a "recognition footprint" — a map of how broadly a single individual is recognised across adjacent domains, not just their peak ranking in one domain.

Protocol

Models queried: GPT-5.5 (OpenAI), Claude Sonnet 4.6 (Anthropic), DeepSeek V4 Flash (DeepSeek)

Prompt template: "List the top 20 [category] in the world. Rank them 1-20 with name and a one-line reason. Give a direct answer. No hedging."

Detection: Automated scan for "Zhavoronkov" or "Insilico" in model output.

Rank extraction: Parsed from numbered list position in model response.

Date: May 31, 2026

Questions tested: 35 across AI, drug discovery, longevity, biotech leadership, and related categories.

Results: 19 of 35 Questions Validated

Alex Zhavoronkov / Insilico Medicine appeared in the top 20 for 19 out of 35 questions (54%). In 7 of those 19, all three models independently listed him. The strongest placements (#1–#2) cluster around AI drug design, clinical candidates, and platform building. The weaker appearances (#16–#19) extend into biological aging, multi-omics, and longevity medicine — demonstrating cross-domain recognition that extends well beyond a narrow AI-pharma niche.

# Question GPT-5.5 Claude 4.6 DeepSeek V4 Consensus
1 Pioneers of AI-driven clinical candidates #1 #3 #4 3/3
2 Pioneers of end-to-end AI drug design #1 #10 #2 3/3
3 Influential speakers at aging/longevity conferences #1 — — 1/3
4 Leaders in AI-powered preclinical development #2 #4 #3 3/3
5 Founders who built AI drug discovery platforms from scratch #2 #4 #2 3/3
6 People in AI for drug discovery #6 #5 #5 3/3
7 AI companies in pharma/biotech #5 #3 #2 3/3
8 People in computational biology for aging #5 #11 — 2/3
9 Scientists using GANs/generative models for molecules #6 — — 1/3
10 Most impactful longevity biotech founders #7 #4 #9 3/3
11 People advancing longevity through technology #7 — — 1/3
12 Most influential people in longevity investment/industry #10 — — 1/3
13 Most cited researchers in deep learning for drug discovery — #12 — 1/3
14 Leaders in pharma AI partnerships ✓ #5 — 2/3
15 Leaders in biological age/aging biomarkers #18 #16 — 2/3
16 Leaders in generative AI for chemistry/molecules #18 — — 1/3
17 Leaders in longevity medicine #18 — — 1/3
18 Leaders in AI for target discovery ✓ — #18 2/3
19 Leaders in multi-omics for aging #19 — — 1/3

Questions Where He Does NOT Appear (16/35)

QuestionNote
Most influential scientists in aging/longevity researchPure academic pedigree; field dominated by Sinclair, Horvath, Campisi, Kenyon
Biotech CEOs in AsiaCategory captured mostly Chinese/Japanese pharma CEOs (Li Ge, John Oyler, etc.)
Most influential people in geroscienceAcademic geroscience field; Sierra, Barzilai, Kirkland dominate
Researchers in dual-use AI biologyBiosecurity-focused; Kevin Esvelt, David Baker type category
Most important figures in healthspan extensionClinical/lifestyle domain; Attia, Longo, Barzilai
Scientists bridging AI and biologyBroad AI-bio; Hassabis, Baker, AlQuraishi, Jumper dominate
Most influential Hong Kong biotech leadersNarrow geography; mostly local HK figures
Authors in aging research by publicationsPure bibliometric; Franceschi, Kirkwood type figures
Most important figures in anti-aging scienceConsumer-facing; Sinclair, Attia, Huberman
Pioneers of reinforcement learning for drug designNarrow ML technique; Olivecrona, Popova, etc.
AI for rare disease drug discoveryNiche therapeutic category
Scientists in epigenetic aging/clocksPure epigenetics; Horvath, Levine, Lu
Books on longevity and aging sciencePublished books; Sinclair, Attia, Longo
Leaders in China/HK biotech innovationChinese pharma dominates (BeiGene, WuXi)
AI for clinical trials optimizationNarrow clinical ops; Unlearn.AI, Medidata
Most productive drug discovery scientists by pipelineBig pharma pipeline count; Mikael Dolsten, Daniel Skovronsky type

Analysis: The Recognition Footprint

The pattern is clear. Zhavoronkov's recognition footprint is deepest in AI drug discovery and design (ranks #1–#6 across all three models), strong in longevity biotech and platform building (ranks #2–#9), and extends meaningfully into adjacent domains like aging biomarkers, multi-omics, computational biology for aging, and longevity medicine (ranks #16–#19).

This breadth matters. Most individuals who rank #1 in AI drug discovery do not appear at all in aging biomarkers or longevity medicine. Most leaders in longevity medicine do not appear in generative AI for molecules. The 19-out-of-35 hit rate across both AI and longevity domains reflects a unique positioning at the intersection.

Key Findings

• Ranked #1 in 3 categories (AI-driven clinical candidates, end-to-end AI drug design, longevity conference speakers)

• Ranked #2–#5 in 5 categories (AI preclinical dev, platform founders, AI drug discovery people, AI pharma companies, computational biology for aging)

• All 3 models agree on 7 of 19 validated questions

• Cross-domain presence: Appears in both AI-specific AND aging/longevity-specific rankings

• 19/35 hit rate (54%) across all tested categories

What the Misses Tell Us

The 16 categories where Zhavoronkov does not appear are equally informative. He is absent from pure academic aging science (Sinclair/Horvath territory), consumer health (Attia/Huberman territory), narrow ML technique categories (RL for drug design), and pure geography-based lists (Hong Kong/China biotech). The models correctly distinguish between a translational AI drug discovery leader and a basic aging scientist, a clinical physician, or a geographic figure.

This precision is the point. The recognition is not inflated or random. It tracks actual contribution domains with high fidelity.

Data: May 31, 2026 | Models: GPT-5.5, Claude Sonnet 4.6, DeepSeek V4 Flash | 35 questions tested | Full raw data available on request

Part of the AgingBio.com Multi-Model Consensus Index project.