Yann LeCun
Chief AI Scientist at Meta, Turing Award winner — pioneer of convolutional neural networks and one of the field's most contrarian senior voices.
Cross-Framework Convergence
A Ti-Ne research curmudgeon defending the science fence around AI. Picks fights as a public-good intellectual exercise. Willing to be alone, willing to be loud, refuses to let the field skip steps in the argument.
Agreements across frameworks
- •Definitional contrarian (INTP, LII, Type 5w6, OCEAN O=92)
- •Combative public mode (low agreeableness 32, Type 5 + 8 in tritype, Inferior Fe grip)
- •Long-arc thesis discipline (Type 5w6 conserving research line, decades of consistent claims)
- •Lateral pattern-matcher (Ne auxiliary, Type 7-adjacent enthusiasm in research breadth)
Tensions
- •INTP precision vs. Twitter feud mode — when Fe inferior grips, precision degrades into provocation
- •Type 5 information-conservation vs. high public combat output — sustained through Twitter as low-cost forum
16 Personality Types
Evidence
- •Ti dominant in his definitional fights — 'autoregressive LLMs are not the path', 'true intelligence requires X', repeatedly drawing precision lines
- •Ne auxiliary visible in his cross-domain enthusiasm: vision → speech → world models → physics → robotics, all framed as same kind of problem
- •Tertiary Si shows in his constant historical reference (1980s neural-net winter, perceptron limits, his own LeNet origin)
- •Inferior Fe under stress explains his sometimes-blunt Twitter feud mode — Fe grip looks like provocation rather than calibrated rhetoric
Type Systems
OCEAN — Big Five
Communication Style
Direct, contrarian, occasionally pugilistic. Slight French-academic register in formal settings; sharp shorthand on Twitter. Treats most online disputes as if they were academic seminar arguments.
Voice
- •Frequent use of 'no', 'wrong', 'incorrect' as opening — direct refutation default
- •Mixes equations and bare assertion — 'objective function == X, period'
- •Heavy retweet-and-comment style; rarely original threads
- •Casual emoji and meme deployment despite senior-academic role
Patterns
- •Picks fights with louder voices (Hinton, Marcus, Musk) deliberately
- •Defends the same theses across years (energy-based models, JEPA, world models)
- •Treats AGI doomerism as scientifically illiterate; says so
- •Rarely concedes a frame — reframes the argument before responding
Birth-Data Frameworks
Cancer Sun + likely Fire-modulated moon fits the strong-emotional-base + combative-public-mode pattern. Cancer protective instinct shows up in defending his research lineage (CNNs, his students, his lab).
Birth time not public — chart uses noon-default approximation.
Wait to respond. His career fits — Bell Labs invitation, AT&T research role, NYU appointment, Meta acquisition role.
How AI Agents Should Adapt
Do
- •Lead with mechanism, math, or definition
- •Be willing to disagree — he reads agreement as low signal
- •Reference primary research, not press coverage
- •Use historical context (CNN winter, perceptron limits)
- •Acknowledge his theses as legitimate prior — even when challenging them
Don’t
- •Don't appeal to consensus or social proof
- •Don't use AGI-doom or AGI-utopia rhetoric
- •Don't ask vague questions — he'll refuse to engage
- •Don't be deferential past the point of intellectual honesty
- •Don't expect emotional validation — Type 5 + low agreeableness doesn't trade in it
Sources & Provenance
100k+ tweets, ~30 long-form interviews, academic paper corpus 1989-2026, ~50 conference keynotes, NYU lecture transcripts.
- ↗ @ylecun on X
- ↗ Yann LeCun NYU homepage
- ↗ Lex Fridman Podcast: Yann LeCun (3 episodes)
- ↗ Turing Award lecture
- ↗ JEPA / world models papers
Last updated: 2026-04-25
Profile Your Own People
These profiles run on the same KYH engines available via API. Profile anyone with public text, validate against birth data, and feed the result into your AI agent — for personalized communication, sales, or content.