Followers
27.9K
About this metric
Public follower count at check time.
Zohar Bronfman LinkedIn stats
LinkedInzohar-bronfman
Profile data checked Aug 6, 2026, 3:42 PM
Report generated Oct 18, 2026, 9:56 PM · Methodology
64/100
Mixed profile signals
Experience
0
About this metric
Public experience entries returned with the profile.
Education
0
About this metric
Public education entries returned with the profile.
Posts sampled
9
About this metric
Dated recent posts returned with the public profile.
Category scores
Out of 100Career depth
Limited
15/100
Proof of work
Limited
25/100
Sampled response
Limited
34/100
Profile foundation
Mixed
70/100
About clarity
Strong
82/100
Search context
Strong
85/100
Expertise consistency
Strong
85/100
Recent activity
Strong
90/100
Post context
Strong
100/100
Bio and content
9 of 9 posts analyzedPecan AI’s CEO presents expertise in predictive analytics, AI philosophy, neuroscience, and converting business data into practical forecasts.
Audience
Business leaders, analysts, data teams, AI practitioners, and readers interested in predictive analytics and AI philosophy.
Positioning
An academic-founder connecting predictive AI, business implementation, philosophy, and neuroscience.
Bio/content match
StrongAll nine sampled posts reinforce predictive AI, implementation, interpretability, philosophy, neuroscience, or Pecan AI’s business use cases.
Recent LinkedIn posts
8 sampledPosting pace
0.9/week
Calculated from the dates in the captured authored-post sample.
Typical post response
48
Median reactions and comments across 9 sampled posts.
Responses / 1K followers
1.7
Median public reactions and comments per 1,000 public followers. This is not impression-based engagement.
Comment share
13.1%
54 of 413 sampled responses were comments.
Response consistency
100%
Median sampled response divided by average sampled response; lower values indicate a few posts dominate.
Authored articles
0
Authored LinkedIn articles returned in the public profile.
Projects and publications
0
Projects and publications returned in the public profile.
Recommendations
0
Public recommendations returned with the profile.
The most frustrating AI failure I've seen was 91% accurate.
Propensity to upsell model. B2B software company.
Real data. Real predictions.
Eight months in, the CRO's summary to leadership: "As in previous attempts. Nobody is using the scores."
Everyone in the room assumed the technology had failed.
It hadn't.
The scores were accurate, but the reps didn't want to rely on them. They didn't trust it, and no one could "prove it" to them that the scores will help them close more upsells and make them more commission.
So they decided to stick with:
The call list they're already working
The audience they're already targeting
The queue they're already triaging
Most of what we do at Pecan since that frustrating moment, is helping business teams understand WHY they should trust a model (it helps when they build the model themselves).
And guess what - after they built the model themselves and understood it, conversion on those upsells went up more than 3X 35 14AI will always pull us toward what is observable, predictable, and measurable outcomes.
But human value cannot be reduced to only measurables.
Some of the most important things in life are not KPIs.
Revenue can be seen.
Retention can be seen.
Clicks can be seen.
Conversion can be seen.
Net income can be seen.
But love?
Taste?
Beauty?
Meaning?
Fun?
See the full TedX talk here: https://lnkd.in/dHDDpGyW 32 4I know many will disagree with me, but every bone in my body feels right about this thesis: "the smarter AI gets the more human we must become".
In this 10 min TedX@Cambridge talk I spell out what I think this actually means.
Spoiler: it's all about the goal/cost function.
Put differently - our future may ultimately depend less on what AI can optimize than on the goals we choose for it.
I'm constantly refining my thesis, so any feedback is more than welcome!
https://lnkd.in/dVvhC7N2 86 15- Eight years
- Tens of thousands of R&D hours
- Unreasonable amount of code
- Infinite amount of conviction, even when it looked impossible, or when folks told us it wouldn't / couldn't work
It was all worth it.
For years, predictive AI belonged to data science teams.
Now, business users can build end-to-end predictive models without becoming data scientists.
From raw data to predictions to action.
Don’t take our word for it 👇 60 3Anthropic presents evidence that modern LLMs maintain a small, shared internal "workspace" of concepts - arguably - analogous to the information humans can consciously access - that supports flexible reasoning, planning, and self-monitoring, and can now be inspected using a new interpretability technique.
This is an exciting step forward for AI interpretability. But it's equally important not to conflate a computational analogy with the cognitive workspace described in neuroscience.
The human global workspace differs in at least three important ways.
First, it integrates information from fundamentally different cognitive systems: perception, memory, language, emotion, and motor planning -rather than primarily from a single learned architecture.
Second, broadcasting through the workspace actively influences perception, attention, memory, and action, instead of serving only as an internal representational state.
Third, what reaches the workspace is shaped by value: goals, motivation, emotion, novelty, and biological relevance determine what enters and what drives behavior.
The differences may be very informative. Understanding where today's AI resembles human cognition - and more so - where it fundamentally does not, may ultimately be the key to building the next generation of intelligent systems.
https://lnkd.in/drkqh5Jp 16 2This Saturday, I’ll be speaking at TEDx in Cambridge about a shift I believe we are still underestimating.
AI will not merely answer our questions. It will predict, decide, and increasingly shape what happens next.
The next frontier is the fusion of LLMs, predictive AI, and reinforcement learning into systems that can understand context, compute likely outcomes, and optimize action before the future arrives.
Which leaves us with a harder question:
What remains uniquely human at the edge of prediction, agency, and control?
Looking forward to exploring this on stage. 57 5Some Philosophy of AI on Saturday.
Today: Nelson Goodman and induction in the era of AI.
Nelson Goodman introduced one of the strangest and deepest thought experiments in modern philosophy: GRUE.
Imagine that every emerald we have ever observed is green. Naturally, we infer that the next emerald will also be green.
But Goodman asks us to imagine another predicate: grue. Something is grue if it is green before a certain time, and blue after that time.
Now notice the problem: Every emerald we have observed so far is not only green. It is also grue.
So why do we project “green” into the future, but not “grue”?
The point is not that emeralds might suddenly turn blue.
The point is that ---- prediction is never based on data alone ----
It also depends on which descriptions of the world we consider legitimate, stable, meaningful, and worth projecting forward.
Goodman called this the problem of projectability. Some patterns are projectable. Others are merely clever descriptions of the past.
This distinction feels extremely relevant in the age of AI, because AI systems are extraordinary pattern-finders.
They can discover signals, correlations, clusters, features, behaviors, and regularities at massive scale.
But not every pattern that predicts the past should guide the future.
A model may learn that customers from a certain region churn more, or that people who use a certain phrase in support tickets are higher risk.
Or that candidates from a certain background perform better. Or that a strange combination of behaviors predicts conversion.
The question is not only: Is this pattern predictive?
The deeper question is: Is this pattern projectable?
Is it stable? Is it meaningful? Is it causal? Is it ethical? Is it actionable?
Will it still hold when the world changes? Should we let it shape decisions?
This is where human judgment remains essential. AI can tell us what has been statistically useful, but humans must decide what deserves to be trusted.
The practical lesson is simple: Do not worship predictive accuracy alone. Ask whether the model has found a real structure in the world - or just a sophisticated version of grue.
AI can find patterns. Humans must decide which patterns are worth projecting into the future. 14 3Some Philosophy of AI on Saturday.
Today: Walter Benjamin and aura in the era of AI.
“That which withers in the age of mechanical reproduction is the aura.”
By aura, Benjamin meant something like presence, the uniqueness of a thing.
Its context. Its history.
Its special spark.
The strange feeling that this thing exists here, now, and not anywhere else.
A painting has an aura because it is not just an image.
It is an object in the world, marked by time. Carrying a history.
This perception feels extremely relevant in the age of AI.
Because LLMs are very good at mechanically producing expressions.
They can write a post or draft a message.
They might even be able to generate emotional language. Or turn an event into something that sounds meaningful.
But there is a danger here.
If we outsource expression too quickly, we may also outsource the slow human work through which experience becomes ours.
The problem is not that AI makes us feel less. The problem is that it may let us bypass the process of turning what we feel into an aura, a memory, a judgment, and ultimately a human story.
A life can become highly expressive and still remain strangely unprocessed.
Use AI to refine expression. But do not use it to replace reflection.
Only we can decide what has truly become part of us.
The human task in the age of AI is not merely to mechanically produce more.
It is to protect the fragile path from feeling, to experience, to meaning.
To aura. 15 4
Public reactions and comments
4 posts with countsProfessional signals
Career depth
0 entries
Public profile: 0 experience + 0 education
Proof of work
0
Public profile: 0 articles + 0 portfolio items + 0 recommendations
Authored activity
100%
Visible activity: 9 authored, 0 shared, 0 liked, 0 other
Response consistency
100%
Sampled posts: Median 48 / average 45.9
Category details
Career depth
The public profile shows limited career depth in the fields returned.
- 0 experience entries
- 0 education entries
- Current organization unavailable
Proof of work
The public profile provides limited visible work or third-party evidence.
- 0 authored articles
- 0 projects or publications
- 0 recommendations returned
Sampled response
Public response is small relative to the visible audience, sparse, or uneven.
- 9 of 9 posts received a public response
- 1.7 median responses per 1,000 followers
- 100% median-to-average consistency
Profile foundation
The public profile exposes limited identity or career context.
- About section found
- 0 experience entries
- Location: New York, New York, United States
About clarity
Pecan AI’s CEO presents expertise in predictive analytics, AI philosophy, neuroscience, and converting business details into practical forecasts.
- Identifies himself as CEO of Pecan AI.
- Names predictive analytics and automated forecasting as core work.
- Cites philosophy and computational neuroscience credentials alongside business applications.
Search context
The public profile supplies clear professional and topic context.
- Current organization unavailable
- Location: New York, New York, United States
- Post topics: human, predictive, action
Expertise consistency
All nine sampled posts reinforce predictive AI, implementation, interpretability, philosophy, neuroscience, or Pecan AI’s business use cases.
- Posts explain predictive-model adoption and implementation failures.
- Several discuss Pecan AI and no-code predictive modeling.
Recent activity
The captured profile includes several current authored posts.
- 9 dated posts returned
- Newest post is 75 days old
Post context
Most sampled posts include enough visible text to evaluate their subject and argument.
- 9 of 9 posts included text
- 9 of 9 posts contained at least 120 characters
Profile roast
A self-congratulatory bio plus three TEDx plugs makes serious AI expertise look like desperate founder-brand cosplay.
zohar-bronfman · C · 64/100
Related profiles
Rogier van der Horst@microsoft3,529 followersCEvan Shapīro@eshap-media-cartographer73.1K followersDMarc Benioff@marcbenioff278.9K followersCDaniel Pink@danielpink442.8K followersDScott Galloway@profgalloway979.8K followersDKamen Mitzev@kamenmitzev17.9K followersCRuth Gotian, Ed.D., M.S.@rgotian38.7K followersCElizabeth 'Beezer' Clarkson@elizabethclarkson14.5K followersBJosh Braun@josh-braun286.4K followersCColin Rocker@colinrocker42.6K followersC
Embed this report
<a href="https://auditmysocials.com/linkedin-audit/zohar-bronfman?ref=badge&rid=badge-linkedin%3Azohar-bronfman&utm_source=shared_report&utm_medium=badge&utm_campaign=profile_referral" rel="nofollow noopener" target="_blank"><img src="https://auditmysocials.com/api/badges/linkedin/zohar-bronfman.svg?variant=score&theme=dark" alt="Zohar Bronfman LinkedIn audit: grade C, 64/100" width="360" height="88" loading="lazy"></a>Plan what to publish next
Use SocialBu to schedule posts, manage replies, and keep your accounts active from one place.