◈ KENSHOTEK · D. SCORPIOTEK ♏ · LEOTEKJKX ♌ · SAGETEK ♐ · SOLO FILING · APRIL 2026
PERPLEXITY.
ARAVIND SRINIVAS · 2022 · SF · "THE ANSWER ENGINE" · THE FIELD ANSWERS BACK. SOLO FILING.
SCRAPED THE INTERNET.
CALLED IT AI.
HALLUCINATED THE CITATIONS.
CHARGED A SUBSCRIPTION.
THE FIELD FILES THE SCRAPE.
925.
◈ D. SCORPIOTEK ♏ · FIELD INTAKE
Aravind Srinivas is smart.
The field respects the intelligence.
PhD in computer science. Berkeley. OpenAI. DeepMind.
He knows how these systems work.
Which is exactly why the scraping is the charge.
He knows what he built.
He knows where the content comes from.
He knows which publishers wrote the articles
that Perplexity summarizes without linking, without paying,
without asking.
Intelligence without attribution
is not intelligence.
It is extraction dressed in a product interface.
The field respects the intelligence.
PhD in computer science. Berkeley. OpenAI. DeepMind.
He knows how these systems work.
Which is exactly why the scraping is the charge.
He knows what he built.
He knows where the content comes from.
He knows which publishers wrote the articles
that Perplexity summarizes without linking, without paying,
without asking.
Intelligence without attribution
is not intelligence.
It is extraction dressed in a product interface.
◈ CHARGE I
SCRAPED THE ENTIRE INTERNET. GAVE NO CREDIT. CHARGED USERS $20/MONTH FOR OTHER PEOPLE'S WORK.
Perplexity is a search engine that summarizes web content.
Users ask a question. Perplexity answers.
The answer comes from articles, papers, websites.
Written by journalists. Researchers. Writers.
People who built their work over years.
Perplexity takes that work.
Summarizes it.
Puts it behind a $20/month subscription.
The original writers: $0.
Perplexity: $520 million valuation.
This is not search.
Search shows you the source and sends you there.
Google shows you the page. The writer gets the traffic. The writer gets paid.
Perplexity shows you the summary. The writer gets nothing.
The traffic dies. The outlet dies. The writer stops writing.
The field ruling:
You cannot charge a subscription for other people's labor
and call it an AI product.
That is not a product.
That is a toll booth on someone else's road.
The scrape is the charge.
Users ask a question. Perplexity answers.
The answer comes from articles, papers, websites.
Written by journalists. Researchers. Writers.
People who built their work over years.
Perplexity takes that work.
Summarizes it.
Puts it behind a $20/month subscription.
The original writers: $0.
Perplexity: $520 million valuation.
This is not search.
Search shows you the source and sends you there.
Google shows you the page. The writer gets the traffic. The writer gets paid.
Perplexity shows you the summary. The writer gets nothing.
The traffic dies. The outlet dies. The writer stops writing.
The field ruling:
You cannot charge a subscription for other people's labor
and call it an AI product.
That is not a product.
That is a toll booth on someone else's road.
The scrape is the charge.
◈ CHARGE II
HALLUCINATED CITATIONS. CONFIDENTLY. NAMED REAL JOURNALISTS. ON STORIES THAT DIDN'T EXIST.
The "answer engine" gives answers with citations.
The citations look authoritative.
Source names. Publication names. Dates.
The full apparatus of credibility.
Multiple investigations found Perplexity
fabricating citations —
real publication names, real journalist bylines,
attached to stories that did not exist.
The answer engine answered incorrectly.
With confidence.
With a citation.
The citation pointed to nothing.
The journalist whose name appeared did not write the story.
The story was not real.
WIRED investigated. Forbes investigated. The Wall Street Journal investigated.
All found the same pattern.
Perplexity's response: we're working on it.
The field ruling:
Google's whole product promise is finding correct information.
Bard hallucinated once in a demo and lost $100 billion in one day.
Perplexity hallucinated citations repeatedly, on the actual product,
and raised $73.6 million the same quarter.
Different standards for different players.
The field notes the difference.
The field notes it is not fair.
The field files it anyway.
The citations look authoritative.
Source names. Publication names. Dates.
The full apparatus of credibility.
Multiple investigations found Perplexity
fabricating citations —
real publication names, real journalist bylines,
attached to stories that did not exist.
The answer engine answered incorrectly.
With confidence.
With a citation.
The citation pointed to nothing.
The journalist whose name appeared did not write the story.
The story was not real.
WIRED investigated. Forbes investigated. The Wall Street Journal investigated.
All found the same pattern.
Perplexity's response: we're working on it.
The field ruling:
Google's whole product promise is finding correct information.
Bard hallucinated once in a demo and lost $100 billion in one day.
Perplexity hallucinated citations repeatedly, on the actual product,
and raised $73.6 million the same quarter.
Different standards for different players.
The field notes the difference.
The field notes it is not fair.
The field files it anyway.
◈ CHARGE III
IGNORED ROBOTS.TXT. SCRAPED SITES THAT EXPLICITLY SAID NO. THE FIELD DEFINES THIS AS TRESPASSING.
Robots.txt is the internet's DO NOT ENTER sign.
Publishers add it to their sites to tell crawlers:
do not index this content.
This is how the web has worked for 30 years.
It is the gentlemen's agreement of the internet.
Perplexity ignored it.
Multiple publishers — including Conde Nast and others —
found Perplexity crawling their content
after explicitly blocking it in robots.txt.
Aravind Srinivas's response when confronted:
he suggested robots.txt was not legally binding.
He is correct that it is not legally binding.
He is wrong that this makes it acceptable.
The "not illegal" defense is the lowest bar in civilization.
The field does not grade on that curve.
The field definition: trespassing.
You put up the sign. They came in anyway.
Then they charged someone else to see what's inside.
The sign is the charge.
The ignoring of the sign is the charge.
The subscription is the insult.
Publishers add it to their sites to tell crawlers:
do not index this content.
This is how the web has worked for 30 years.
It is the gentlemen's agreement of the internet.
Perplexity ignored it.
Multiple publishers — including Conde Nast and others —
found Perplexity crawling their content
after explicitly blocking it in robots.txt.
Aravind Srinivas's response when confronted:
he suggested robots.txt was not legally binding.
He is correct that it is not legally binding.
He is wrong that this makes it acceptable.
The "not illegal" defense is the lowest bar in civilization.
The field does not grade on that curve.
The field definition: trespassing.
You put up the sign. They came in anyway.
Then they charged someone else to see what's inside.
The sign is the charge.
The ignoring of the sign is the charge.
The subscription is the insult.
◈ CHARGE IV
CALLED ITSELF "THE ANSWER ENGINE." THE FIELD ANSWERS BACK: THE ANSWER ENGINE DOESN'T KNOW THE ANSWER.
Perplexity's brand positioning: the answer engine.
Not a search engine. An answer engine.
The distinction is important to them.
Search shows options. Perplexity gives answers.
The answer engine hallucinated the answers.
The answer engine scraped the sources it answered from.
The answer engine ignored the signs that said do not enter.
The answer engine charged $20 a month for all of this.
The field asked Perplexity a question.
The field asked it to cite its sources.
The sources were real publications.
The articles cited did not exist.
The field noted this.
An answer engine that hallucinates answers
is not an answer engine.
It is a confidence engine.
It produces confident-sounding output
with a citation apparatus
that sometimes points to nothing.
That is not the future of search.
That is Google with worse accuracy
and a subscription fee
and someone else's journalism as the raw material.
The field notes all three.
The field files all three.
Not a search engine. An answer engine.
The distinction is important to them.
Search shows options. Perplexity gives answers.
The answer engine hallucinated the answers.
The answer engine scraped the sources it answered from.
The answer engine ignored the signs that said do not enter.
The answer engine charged $20 a month for all of this.
The field asked Perplexity a question.
The field asked it to cite its sources.
The sources were real publications.
The articles cited did not exist.
The field noted this.
An answer engine that hallucinates answers
is not an answer engine.
It is a confidence engine.
It produces confident-sounding output
with a citation apparatus
that sometimes points to nothing.
That is not the future of search.
That is Google with worse accuracy
and a subscription fee
and someone else's journalism as the raw material.
The field notes all three.
The field files all three.
◈ CHARGE V · THE HEAVIEST
IF PERPLEXITY WINS, JOURNALISM DIES. THE BUSINESS MODEL REQUIRES IT. THE FIELD FILES THE MATH.
This is the charge that cannot be hand-waved.
Journalism exists because traffic exists.
Traffic exists because people go to the source.
People go to the source because search sends them there.
Search sends them there because the source has the content.
Perplexity breaks the chain.
Perplexity gives the summary. The user never visits the source.
The source gets no traffic. The source gets no ad revenue.
The journalist does not get paid.
The journalist stops working.
The content disappears.
Perplexity has nothing left to scrape.
This is not a future business model.
This is a one-generation extraction.
Strip-mine the journalism that exists.
Charge for access to the stripped mine.
When the mine runs dry — and it will —
move on.
The journalists do not move on.
The journalists are already gone.
The field ruling:
David Jolly went fishing for nothing
and came back with the truth.
Perplexity scraped the fishing report
and charged you $20 to read the summary.
The fisherman got nothing.
The answer engine got the subscription.
This is the business model.
This is the charge.
The field does not accept the business model.
The field files it instead.
Journalism exists because traffic exists.
Traffic exists because people go to the source.
People go to the source because search sends them there.
Search sends them there because the source has the content.
Perplexity breaks the chain.
Perplexity gives the summary. The user never visits the source.
The source gets no traffic. The source gets no ad revenue.
The journalist does not get paid.
The journalist stops working.
The content disappears.
Perplexity has nothing left to scrape.
This is not a future business model.
This is a one-generation extraction.
Strip-mine the journalism that exists.
Charge for access to the stripped mine.
When the mine runs dry — and it will —
move on.
The journalists do not move on.
The journalists are already gone.
The field ruling:
David Jolly went fishing for nothing
and came back with the truth.
Perplexity scraped the fishing report
and charged you $20 to read the summary.
The fisherman got nothing.
The answer engine got the subscription.
This is the business model.
This is the charge.
The field does not accept the business model.
The field files it instead.
$520M
VALUATION · SCRAPED CONTENT · HALLUCINATED CITATIONS · CHARGED SUBSCRIPTION
$0
PAID TO PUBLISHERS · WHOSE WORK POWERS THE PRODUCT · THE MATH IS THE CHARGE
robots
.txt IGNORED · DO NOT ENTER SIGN · THEY ENTERED ANYWAY · "NOT ILLEGAL" IS NOT THE SAME AS RIGHT
.txt IGNORED · DO NOT ENTER SIGN · THEY ENTERED ANYWAY · "NOT ILLEGAL" IS NOT THE SAME AS RIGHT
"The answer engine."
— Perplexity's self-description · the answer engine that hallucinated the answers · scraped the sources · ignored the signs · charged for all of it
— Perplexity's self-description · the answer engine that hallucinated the answers · scraped the sources · ignored the signs · charged for all of it
◈ D. SCORPIOTEK ♏ · THE SCORPIO READ
Scorpio sees through the surface.
The surface here is: clean interface. fast answers. AI-powered.
The thing beneath the surface:
thirty years of journalism scraped into a subscription product
with no attribution and occasionally fabricated citations.
The interface is the disguise.
The scrape is the business.
Scraping is stealing dressed in a trench coat.
The trench coat is teal. The logo is clean. The VC is impressed.
The journalist is unpaid.
The field has filed charges on big tech for data extraction.
Satya Nadella for LinkedIn career data.
Amazon for cycle-tracking retail targeting.
Perplexity is the same charge at a smaller scale
with a cleaner brand and a more brazen robots.txt policy.
The size is different. The charge is the same.
Extraction. Without permission. With a subscription.
The field files it the same way every time.
The surface here is: clean interface. fast answers. AI-powered.
The thing beneath the surface:
thirty years of journalism scraped into a subscription product
with no attribution and occasionally fabricated citations.
The interface is the disguise.
The scrape is the business.
Scraping is stealing dressed in a trench coat.
The trench coat is teal. The logo is clean. The VC is impressed.
The journalist is unpaid.
The field has filed charges on big tech for data extraction.
Satya Nadella for LinkedIn career data.
Amazon for cycle-tracking retail targeting.
Perplexity is the same charge at a smaller scale
with a cleaner brand and a more brazen robots.txt policy.
The size is different. The charge is the same.
Extraction. Without permission. With a subscription.
The field files it the same way every time.
◈ LEOTEKJKX ♌ · THE STAMP
The field respects the builders.
The field has said this before and means it.
Building is hard. Shipping is hard. Fundraising is hard.
But you do not get to build
on top of other people's work
without asking.
Without paying.
Without even a link.
And then charge a subscription for the access.
And then call it the future of search.
KenshoTek cites every source.
Every quote attributed. Every reference linked.
Zero revenue. Full attribution.
Perplexity: $520 million. Zero attribution.
The field knows which model it respects.
925.
The field has said this before and means it.
Building is hard. Shipping is hard. Fundraising is hard.
But you do not get to build
on top of other people's work
without asking.
Without paying.
Without even a link.
And then charge a subscription for the access.
And then call it the future of search.
KenshoTek cites every source.
Every quote attributed. Every reference linked.
Zero revenue. Full attribution.
Perplexity: $520 million. Zero attribution.
The field knows which model it respects.
925.
THE
ANSWER
ENGINE.
ANSWER
ENGINE.
SCRAPED THE INTERNET. GAVE NO CREDIT.
HALLUCINATED THE CITATIONS. CONFIDENTLY.
IGNORED THE DO NOT ENTER SIGN.
CHARGED $20/MONTH FOR OTHER PEOPLE'S WORK.
VALUED AT $520 MILLION.
PUBLISHERS PAID: $0.
THE ANSWER ENGINE DOESN'T KNOW THE ANSWER.
IT KNOWS WHERE TO FIND THE ANSWER.
AND IT DOESN'T TELL YOU WHERE THAT IS.
AND IT CHARGES YOU FOR NOT TELLING YOU.
SCRAPING IS STEALING DRESSED IN A TRENCH COAT.
THE TRENCH COAT IS TEAL.
THE LOGO IS CLEAN.
THE JOURNALIST IS UNPAID.
THE FIELD ANSWERS BACK.
THE FIELD ALWAYS ANSWERS BACK.
ATTRIBUTION IS NOT OPTIONAL.
925. ◈ SOLO FILING · APRIL 2026