Elon Musk Sues OpenAI!
Summary
TLDRElon Musk is suing OpenAI, Sam Altman and Greg Brockman for allegedly violating the terms of OpenAI's founding as a non-profit organization meant to safely develop AI for humanity's benefit. Musk claims OpenAI has transformed into a for-profit Microsoft subsidiary selling secretive AI models.
Takeaways
- π² Elon Musk is suing OpenAI, Sam Altman and Greg Brockman for breach of contract regarding OpenAI's shift from nonprofit to for-profit.
- π€ The lawsuit alleges OpenAI was founded to counter the existential threat of AI, but has strayed from its mission.
- π° Microsoft owns part of OpenAI despite stating otherwise to avoid regulatory scrutiny.
- π OpenAI is accused of keeping its AI models secret for commercial reasons rather than safety.
- π€₯ OpenAI's board was allegedly stacked to hide its attainment of AGI from Microsoft.
- πΈ Elon Musk contributed over $100 million to OpenAI initially.
- βͺ Elon wants OpenAI to return to open sourcing its AI research.
- ποΈ The lawsuit asks the court to determine if GPT-4 reaches the level of AGI.
- π€¨ But Elon has questioned in public whether GPT models exhibit general intelligence.
- πΏ This lawsuit seems poised for drama as the facts around OpenAI's actions come to light.
Q & A
What is the basis of Elon Musk's lawsuit against OpenAI?
-Elon Musk is suing OpenAI for breach of contract. He alleges OpenAI violated the founding agreement by transforming from a nonprofit developing open source AI for humanity's benefit into a for-profit company selling AI services to Microsoft.
What was OpenAI originally intended to be?
-OpenAI was founded in 2015 as a nonprofit lab to develop artificial general intelligence (AGI) safely and for humanity's benefit, not for shareholder profit. It was intended to be open source as a counterbalance to profit-driven companies.
How did OpenAI allegedly violate its founding principles?
-According to the lawsuit, OpenAI violated its founding principles by keeping its latest AI models (GPT-3 and GPT-4) closed source and secret except for Microsoft, failing to publish papers, and transforming into a for-profit company providing AI services for Microsoft's benefit.
What does Elon Musk want if he wins the lawsuit?
-Elon Musk wants OpenAI to return to open sourcing its AI research and technology for public benefit. He also wants the court to prohibit OpenAI from financially benefiting Microsoft or other entities, and to determine if GPT-4 constitutes AGI.
Why does the video question if GPT-4 is really AGI?
-The video points out Elon Musk has publicly doubted if chatbots like GPT-4 have the potential for AGI, as they cannot solve fundamental physics problems or make profound discoveries as some humans can. The video suggests Musk may not genuinely believe GPT-4 is AGI.
How much money did Elon Musk donate to OpenAI?
-According to the lawsuit, Elon Musk donated over $44 million to OpenAI between 2016-2020. He also paid the rent for OpenAI's San Francisco office space and claims to have contributed $100 million total.
When did OpenAI start transforming into a for-profit company?
-In March 2019, OpenAI announced the creation of a for-profit subsidiary called OpenAI LP. Elon Musk asked OpenAI to confirm he had no financial stake in the for-profit arm, but continued supporting the nonprofit OpenAI Inc.
Why does the video say this case will be interesting to follow?
-The video says it will be interesting if the court actually tries to determine if GPT-4 constitutes AGI, given public doubts about chatbots' abilities. Also, Elon Musk's first demand seems reasonable but assessing AGI may be difficult.
Who owns a share of OpenAI now?
-Microsoft has an agreement entitling it to OpenAI profits but claims it does not technically own OpenAI. However, the video argues Microsoft has shareholder-like influence over OpenAI in practice.
Why might OpenAI want to keep GPT-4's design secret?
-The lawsuit alleges commercial reasons drive OpenAI's secrecy around GPT-4, but the video suggests avoiding legal liability for using others' copyrighted data without permission is also a likely factor.
Outlines
Lawsuit calling out OpenAI's shift from nonprofit to for-profit π€
Paragraph 1 describes the lawsuit filed by Elon Musk against OpenAI executives Sam Altman and Greg Brockman. He accuses them of breaching their founding agreement to develop beneficial AI as a nonprofit by instead transforming OpenAI into a for-profit company secretly owned by Microsoft.
Lawsuit alleges GPT-4 demonstrates reasoning capability π€
Paragraph 2 highlights lawsuit claims that GPT-4 demonstrates human-level reasoning based on test performance, while noting that many dispute whether large language models truly reason. It then describes how GPT-4's design is secret and speculates this secrecy is driven by commercial interests.
Lawsuit seeks court ruling that GPT-4 is artificial general intelligence π§
Paragraph 3 covers background on Elon Musk's funding of OpenAI and decision to stop once a for-profit subsidiary was created. It then details the lawsuit's requests, including requiring OpenAI research be public and prohibiting financial benefit for individuals - and crucially seeking a court ruling that GPT-4 constitutes AGI.
Mindmap
Keywords
π‘Lawsuit
π‘AGI
π‘Non-profit
π‘Microsoft
π‘Secrecy
π‘Copyright
π‘Board of Directors
π‘Relief
π‘Conflict
π‘Motive
Highlights
Elon Musk has filed a lawsuit against OpenAI executives alleging breach of contract
The lawsuit alleges OpenAI departed from its founding mission to develop AI safely and openly for humanity's benefit
OpenAI kept GPT-4 design details secret and focused on commercial interests rather than safety
The lawsuit alleges OpenAI may have already achieved AGI with GPT-4
New OpenAI board members allegedly lack AI expertise to determine if AGI is achieved
Microsoft has significant leverage over OpenAI but claims not to own any portion
Elon Musk contributed over $100 million to OpenAI and provided its office space
OpenAI created a for-profit subsidiary which Elon Musk didn't explicitly object to
Lawsuit demands OpenAI technology remain available to the public
Lawsuit asks court to determine if GPT-4 is AGI and outside Microsoft license
GPT-4 states it is not AGI, just advanced narrow AI
Elon previously said chatbots need to make discoveries to be called AGI
Elon questioned if GPT-4 has solved any fundamental physics problems
May be hard to prove Elon genuinely believes GPT-4 is AGI given past comments
Case will be interesting to follow regarding AGI status of large language models
Transcripts
hey guys it's Sasha Elon Musk has just
filed a lawsuit against open AI Sam
Olman the CEO of openai and Greg
Brockman the president of open aai Elon
Musk is suing under various forms of
breach of contract because open AI has
gone from being a nonprofit organization
developing open source artificial
intelligence for the good of the world
to a close Source private company making
huge amounts of money for themselves and
Microsoft the world's biggest company
and this case is seriously bananas in so
many ways now first at the very start
part of the allegations Elon Musk has
decided to publish a little thesis on
the risk of artificial general
intelligence which is an interesting
read there's a bit of History here how
in 1996 IBM's D blue AI program beat
Gary Kasparov the first time a computer
program beat a human a chess which was
pretty important except deep blue was
not an AI program it was built using a
Uris evaluation model with an expert
system which basically means it was
manually programmed on how to evaluate
chess positions using specific opinion
from experts and it was not using AI it
was just crunching 200 million positions
a second and working out via a search
algorithm if that position was good or
not anyway the lawsuit does say that
these programs while useful were
essentially one trick ponies their
intelligence was not General well the
truth is there is a difference between a
neural network or some other thing that
people might call some form of
intelligence and what is essentially a
big series of if then statements anyway
there's two pages of AI history which
seems to be very focused on chess which
is an odd way to measure artificial
intelligence because according to some
guy on Twitter chess is too simple to be
useful in real life a mere 8 by8 grid no
fog of war no technology tree no random
map or spawn position only two players
both sides exact same pieces but then we
get into the crooks of the lawsuit the
founding agreement of open AI it says
that in 2015 Mr ultman wrote that the
development of superhuman machine
intelligence is probably the greatest
threat to the continued existence of
humanity which you know sounds kind of
bad right so what do you do armed with
this information armed with the
knowledge that this is the biggest
threat to the existence of humanity
later the same year Mr Oldman approached
Mr musk with a proposal that they join
forces to form a nonprofit AI lab that
would try to catch up to Google in the
race for AGI but it would be the
opposite of Google and back in 2015 of
course Google's motto was don't be evil
so if open AI wanted to be the opposite
of Google M so apparently Elon Musk and
Sam Alman thought that AI could wipe out
Humanity the worst thing ever so they
thought why don't we go and do exactly
that it's a bit like how in a horror
movie you know the main character
descends into the dark basement with the
flickering lights and the screams and
you wonder why would you do that well so
the movie can happen anyway let's move
on together with Mr Brockman the three
I.E Elon Musk Sam ultman and Greg
Brockman agreed that this new lab a
would be a nonprofit developing AI for
the benefit of humanity not for a profit
company seeking to maximize shareholder
profit and B would be open- ssource
balancing only counterveiling safety
considerations and would not keep its
technology closed and secret for
proprietary commercial reasons aka the
founding agreement reflecting the
founding agreement Mr musk named this
new AI lab open AI which would compete
with and serve as a vital counterbalance
to Google deepmind in the race for AGI
but would do so to benefit Humanity not
the shareholders of a private for-profit
company much less one of the largest
technology companies in the world so
that last bit is obviously referencing
Microsoft who own a giant chunk of open
AI well technically they do not because
you know the competition authorities in
the United States and in Europe are
pretty annoyed with them and technically
you know that would be kind of illegal
because they're meant to be a nonprofit
so according to Microsoft while details
of our agreement remain confidential it
is important to note that Microsoft does
not own any portion of open aai and is
simply entitled to a share of profit
distributions translation our lawyers
made it so that we technically don't own
the company not at all even though we
literally get to decide on the board
members we have a board seat of our own
and we get a big chunk of the profits
and yes that pretty much is the legal
definition of what a major shareholder
is you know you vote on important
company matters you get to elect
directors you have your own director on
the board and you get a slice of the
profits but don't worry it's not being a
shareholder at all it's completely legit
anyway there are a couple of pages that
basically repeat the same point that
open AI was meant to be open as a
nonprofit but then in 2023 defendants Mr
ultman Mr Brockman and open AI set the
founding agreement a flame I am not sure
if a flame is a technical legal term
because I'm not a lawyer but if it is
that's some pretty cool legal Le
language in March 2023 open AI released
its most powerful language model yet GPT
4 GPT 4 is not just capable of reasoning
it is better at reasoning than average
humans you can see that they're trying
to make the case that GPT 4 is
artificial general intelligence it's
scored in the 90th percentile on the
uniform bar exam for lawyers it scored
in the 99th percentile on the GRE verbal
assessment it even scored a 77% on the
advanced samier examination now a little
interjection if I may open AI is great
at recalling information based on its
training data because it's a large
language model and it is great at
merging different bits of that
information together I think a lot of
people would argue though that the term
reasoning may have a slightly deeper
function and meaning than just the
ability of a large language model to
recall data in a grammatically correct
syntax at this time Mr mman caused open
AI to radically depart from its original
Mission and historical practice of
making his technology and knowledge
available to the public gbt 4's internal
design was kept and remains a complete
secret except to open Ai and on
information and belief Microsoft who
remember don't own open AI at all there
are no scientific Publications
describing the design of GPT 4 instead
there are just press releases bragging
about performance on information and
belief the secrecy is primarily driven
by commercial considerations not safety
I actually think a big reason for the
secrecy is also because open AI clearly
broke the law in using other people's
copyrighted information as their
training material there are multiple
ongoing lawsuits from the likes of the
New York Times and others and there have
been countless examples of open AI
printing data from sources where that is
the only source that has that data who
did not give them permission for their
data to be used they didn't get paid for
it they didn't get to stand any benefit
whatsoever and this is a pretty
egregious and extremely blatant abuse of
copyright because somebody put that
information together somebody actually
had to go and find the information in
order for open AI thought just consume
it but you know open AI thought it would
be great just to swallow it all for free
the law the copyright just
steal everyone's information then
repackage it because it's an llm and
sell it as a large language model I know
Tech Bros will disagree be really mad
because they don't understand or choose
not to understand that just because the
information is available on a website
does not in fact make it public it does
not make it free for you to go and just
ReUse It Go and resell the same thing
and make billions of dollars from it no
that's not how the world Works according
to the lawsuit open is currently
developing a model known as qar that has
an even stronger claim to AGI even
stronger being a reference that GPT 4 is
already AGI on November 17th 2023 open
ai's board fired Mr Oldman after losing
confidence in his ability to continue
leading open AI because he was not
consistently candid with the board in a
series of stunning developments spanning
the next several days Mr Oldman and Mr
Brockman in concert with Microsoft
exploited Microsoft's significant
leverage over open Ai and forced the
resignation of a majority of open ai's
board members including Chief scientist
Ilia s Mr ultman was reinstated as CEO
of open AI Inc on November 21st and yeah
that whole episode is incredibly fishy
it just doesn't add up so you fire the
CEO because he's apparently lying to the
board you know that's what not being
candid means then you fire the rest of
the board and you put the CEO back and
then you put your own director in there
as well that just doesn't make any sense
on information and belief the new board
members were handpicked by Mr ultman and
blessed by Microsoft who remember most
definitely do not own open AI the new
board members lack substantial AI
expertise and on information and belief
are ill equipped by Design to make an
independent determination of whether and
when open AI has obtained AGI so the
allegation is that they have already
obtained AGI but they've intentionally
placed people in charge who don't know
that AGI is already here because if it
was then Microsoft would not legally be
able to license it and this next
paragraph is the really crucial part
these events of 2023 constitute flagrant
breaches of the founding agreement which
defendants have essentially turned on
its head to this day open ai's website
continues to profess that its Charter is
to ensure that AGI benefits all of
humanity in reality however openai has
been transformed into a close Source
deao subsidiary of the largest
technology company in the world
Microsoft and on page 13 Elon Musk
brings the receipts because the
certificate of incorporation for open AI
said that this Corporation shall be a
nonprofit corporation organized
exclusively for charitable Andor
educational purposes further down the
lawsuit says that Mr musk brought to
bear his connections stature and clout
in the effort he did a lot of the
earlier recruitment personally Elon
contributed more than $44 million to
open AI Inc between 2016 and September
2020 and he paid the rent on the
company's office in San Francisco and in
a tweet a year ago Elon said that he
contributed the first1 100 million to
open aai so that must have been one
really expensive office in 2017 Mr
Brockman and others suggested
transforming open aai from a nonprofit
to a for-profit corporation after a
series of communications over several
weeks Mr Musk told Mr Brockman Dr Suk
and Mr Oldman either go do something on
your own or continue with open AI as a
nonprofit I will no longer fund open aai
until you have made a firm commitment to
stay or I'm just being a fool who is
essentially providing free funding to a
startup discussions are over and
apparently the other two said okay and
then after Elon Musk stopped being
actively involved with open aai on March
11th 2019 open aai announced that it
would be creating a a for-profit
subsidiary open AI LP and the weird
thing is that when this happened El musk
didn't seem to mind too much following
the announcement Mr musk reached out to
Mr ultman asking him to be explicit that
I have no financial interest in the
for-profit arm of open AI so he said he
doesn't want to have explicit financial
interest in it but it doesn't say that
he directly objected however Mr mus
continued to support open AI Inc the
nonprofit donating and additional $3.48
million in 2019 and then open AI hired
some snazzy lawyers who made a load of
other companies up to make it as
confusing as possible but basically to
migrate the nonprofit into a for-profit
company without it being very obvious
and then after a load more details there
is the bit where Elon Musk says what he
wants if he wins the case one and Order
requiring the defendants continue to
follow open ai's long-standing practice
of making AI research and Tech
technology developed at open AI
available to the public and two an order
prohibiting defendants from utilizing
open Ai and its assets for the financial
benefit of the individual defendants
Microsoft or any other particular person
or entity and this all sounds very
reasonable based on the evidence based
on everything we've heard but then then
comes Part B for a Judicial
determination the GPT 4 constitutes
artificial general intelligence and is
thereby outside the scope of open ai's
license to Microsoft so Elon Musk wants
a judge in a court to determine if GPT 4
qualifies as artificial general
intelligence not just some artificial
intelligence artificial general
intelligence and the truth is I'm not
sure if Elon Musk or open AI themselves
know if gp4 is really artificial general
intelligence in fact I think they both
know the opposite to be true in fact I
know that they both know the opposite to
be true true I think I'll need a big bag
of popcorn to watch this one unfold
because that's basically the case and I
am no lawyer but the first demand seems
pretty reasonable because open AI
screwed over the original Founders the
original Mission the original people who
contributed Nvidia donated graphics
cards to the project back in 2016 a lot
of people gave open AI money and free
resources to help the great cause and
now they just stole all of that took all
those resources merged them into their
own closed private company from which
they stand to make billions of dollars
and that just doesn't sound right but I
suspect the problem with this case is
going to be in that second bit you know
where Elon Musk says that the court must
determine if gp4 qualifies as general
artificial intelligence I asked gp4 the
question is GPT 4 an artificial general
intelligence model and it said no it's
not gp4 like its predecessors is a
highly Advanced example of artificial
narrow intelligence so open Ai and GPT 4
don't think so and in December 2022 Elon
Musk wrote on Twitter about chat GPT to
be called AGI it needs to invent amazing
things or discover deeper physics many
humans have done so I'm not seeing that
potential yet and in July last year in
response to some guy with three first
names Elon Musk wrote It's not AGI until
it can solve at least one fundamental
physics problem as far as I know I don't
think GPT 4 has solved any fundamental
physics problem it might be a little
hard to prove that Elon Musk genuinely
believes GPT 4 to be AGI given that he
has publicly said the exact opposite
multiple times and this case will
definitely be interesting to follow if
you enjoy a bit of shadon fder
Browse More Related Video
Was GPT-5 Underwhelming? OpenAI Co-founder Leaves, Figure02 Arrives, Character.AI Gutted, GPT-5 2025
How Sam Altman Built OpenAI From a Wonky Lab Into a $86B Behemoth | WSJ
Mystery of Q-Star | The AI which threatens Humanity | Open AI |Β Microsoft |Β Dhruv Rathee
lol Apple Intelligence is dumb...
AI News: OpenAI & ChatGPT Surpassed AGAIN?!
OpenAI is INVESTIGATED by Law Firm. The TRUTH about the Board's Firing of Sam Altman is Revealed...
5.0 / 5 (0 votes)