Elon Musk names OpenAI | Mark Zuckerberg | Bill Gates
Welcome to Internal Tech Emails: internal tech industry emails that surface in public records. đ If you havenât signed up, join 50,000+ others and get the newsletter:
Mark Zuckerberg: âDo you think itâs way too much?â
April 5, 2012
Mark Zuckerberg
I think we're going to get Instagram done at 1.3% or 1.4%. I'm breaking to interview someone now but Kevin and I will likely agree on a number later tonight.
Can you get the team mobilized to get together and get this done quickly? He has extreme time pressure because of his other investment, so if we can get this done Saturday that is better than Sunday. They should have all day tomorrow to work.
Sheryl Sandberg
Yes!
Sheryl Sandberg
starting an email thread to get started --we can move fast. Please respond so we get all of the info we need.
Mark Zuckerberg
He's calling Cohler now to get approval. Cohler doesn't want to do it since he thinks price is too low.
Sheryl Sandberg
One other thing, Amin was [REDACTED] boss - he worked for him in corp dev. So this is good.
Mark Zuckerberg
It will be ~1.38% of our stock.
Mark Zuckerberg
[REDACTED]? You mean Kevin?
Sheryl Sandberg
yes - sorry - tired
Sheryl Sandberg
But, Amin thinks we are dramatically overpaying. I talked him through it but he wants to talk to you. Please take 10 min to talk to him in the AM. respectful and he rocks.
Sheryl Sandberg
he will do the deal as if he believes - but just a nice thing to do
Mark Zuckerberg
Just want to gut check that this is reasonable. Do you think it's way too much?
Sheryl Sandberg
can I call you?
Sheryl Sandberg
yes of course it is way too much
Sheryl Sandberg
but we knew that
Sheryl Sandberg
calling
Sheryl Sandberg
just spoke to [REDACTED]
Sheryl Sandberg
calling you now
Mark Zuckerberg
Kevin came back and said he's not sure if he's willing to push his investors to do this. I think he likely means it. He's going to sleep on it and we're going to meet up in the morning. I'll let you know how that goes.
[This document is from FTC v. Meta (2025).]
Previously: Mark Zuckerberg: "It's a really big deal that we ship this photos appâ (September 11, 2011)
Previously: Instagram cofounder on Mark Zuckerberg: âwill he go into destroy mode if I say noâ (February 13, 2012)
Previously: Mark Zuckerberg tries to buy Instagram (Circa March 2012)
Previously: Mark Zuckerberg: Should we buy Instagram, Foursquare, or Pinterest? (April 5, 2012)
Previously: Mark Zuckerberg: âlast night we reached an agreement to acquire Instagramâ (April 9, 2012)
A note from @TechEmails
Every year, I track hundreds of court cases and review more than 10,000 filings to bring you Internal Tech Emails. If you like @TechEmails, and would like to help make this work more sustainable, consider upgrading to a paid subscription.
Youâll be supporting the research that drives Internal Tech Emails, and will help ensure that it can continue publishing. And youâll also receive access to the full archive of internal tech emails, with 200+ documents from Apple, Google, Meta, Microsoft, OpenAI, Tesla, and more.
Thanks for reading!
-Internal Tech Emails
Sent from my iPad
Greg Brockman and Ilya Sutskever on naming OpenAI
November 24, 2015
Greg Brockman
What do you think Cogito as a name?
Ilya Sutskever
Not a huge fan
Greg Brockman
noooooo!
(elon + sam both liked it)
Ilya Sutskever
What about you? :)
Greg Brockman
i came up with it âș
Ilya Sutskever
Oops!
Greg Brockman
haha
but, i like that it's relevant, pretty unbranded, and has good connotations of thinking/individualism
i don't like that it's hard to know how to pronounce.
Ilya Sutskever
Ah, is it Latin?
Greg Brockman
ah yeah. "cogito ergo sum"
Ilya Sutskever
Ah
Greg Brockman
descartes' "i think therefore i am"
Ilya Sutskever
I can see why everyone liked it
I had friends who named their company "whetlab"I didn't like it at once since I didn't know what whet meant
If you think that enough people know what cogito means than i support it
Greg Brockman
elon says:
Not bad. Sounds kinda cute. Most people won't get the latin, but the ones we want to join will.
I'd support that.
(or will be russian âș)
Ilya Sutskever
Knowledge of Latin is independent of knowledge of ML
Greg Brockman
hah, yeah. but more seriously, i think cogito is well-known enough
Ilya Sutskever
Maybe it's ok
Then I'm ok with it
Greg Brockman
cool
[This document is from OpenAI v. Open Artificial Intelligence (2025).]
Elon Musk on naming OpenAI
On Dec 2, 2015, Greg Brockman wrote:
One update from me:
On name, there's a potentially conflicting trademark on Cogito. What do you think of "Consider" (we can get consider.com and @consider for a reasonable price)? I like that it's otherwise unbranded, has no negative connotations, and also points to our differentiator: we consider the impact of AGI.
On Thu, Dec 3, 2015, Elon Musk wrote:
Don't love the sound of Consider. What's the Cogito trademark issue?
A naming approach we could try is for the real name to be long, but have actual use be a contraction, eg the full name of SpaceX is Space Exploration Technologies Corp.
We could call this the Open Al Institute (Open AI-I!), but call it Open in everyday conversation. Looks like OpenAI.com is available for purchase.
On Dec 3, 2015, Greg Brockman wrote:
The relevant trademark is here: http://www.trademarkia.com/cogito-79117699.html (it's for this product: http://www.expertsystem.com/cogito/).
And aww re Consider. It's grown on me a lot as I've tried it out in various contexts, and think we could make it work pretty well. (I had a similar experience with the name "Stripe". At choosing time it felt pretty random, but because it's otherwise so unbranded it started to feel more natural as time went on.) How down on it are you?
On Thu, Dec 3, 2015, Elon Musk wrote:
Pretty down. Consider sounds a bit nannyish and self-righteous.
On Thu, Dec 3, 2015, Greg Brockman wrote:
Ok, got it.
Looks like there are a few things called OpenAI (https://openai.org/, http://openai.sourceforge.net/), though nothing really established. I could get behind it if we can't find anything else we like. ThinkingâŠ
FWIW, on the short name front, here's a list of domains a friend-of-friend is selling/brokering: [REDACTED].
Nothing hugely stood out to me besides Consider though.
On Dec 3, 2015, Greg Brockman wrote:
Ok, I'm actually starting to feel pretty good about OpenAI. Also polled a few people I trust, and all good reactions. I've called the owner of openai.com; working on him now.
Branding-wise, I'm thinking we can call it "OpenAI" or "the OpenAI group". (Because our ultimate goal is the combined engineering and research goal of AGI, I think we should avoid "institute" or "research institute" in our positioning.) I could be convinced otherwise though.
On Thu, Dec 3, 2015, Elon Musk wrote:
Yeah, good point about not sounding too researching. I mostly liked Institute, because it sounded like ayiyi as an acronym :) Just OpenAI is probably better.
[This document is from OpenAI v. Open Artificial Intelligence (2025).]
Previously: Sam Altman emails Elon Musk: âIf [AI is] going to happen anyway, it seems like it would be good for someone other than Google to do it firstâ (May 25, 2015)
Previously: Sam Altman emails Elon Musk: âThe mission would be to create the first general Alâ (June 24, 2015)
Previously: Ilya Sutskever emails Elon Musk: âYou are concerned that Demis could create an AGI dictatorship. So do weâ (September 20, 2017)
Bill Gates emails Steve Ballmer
From: billg
To: steveb
Subject: OS strategy
Date: Sun Aug 6 21:34:03 1989
I have been thinking about this in a pure MS context to understand what is important for us and then see how the IBM relationship helps with that.
ny worth $7B. Thats doubling the current valuation. We would have to more than double our profits. Let say systems has to bear the same proportion of that burden in the future as it does now and that as a group (server, mouse, lang, and toolkits) do their share (small) with server making up for the fact that mouse, lang and tollkits cant bear their part. That leaves the Os part (dos, win, os/2) with a huge burden. Basicly last year (I may get the numbers wrong) it was around $180M worldwide PROFIT. Assuming 12m machines thats $15 PROFIT per machine or giving IBM 1/3 of the market (a little high) $22.5 per non IBM machine. We have to drive that number up to around $45. Taking a 3 year time frame I think windows has a lot to do with it âI can see us lowering the price differential for "high end windows" (a nickname I have for what os/2 has to be perceived as over time) because we want to get it bootstrapped to the point where it wont cost a lot more than windows does. This means the price we can charge for windows will drive a lot of our future. Now how can we get this doubling per system thing to take place.
First, we have to make sure windows isnt easy to clone for both technical and legal reasons. Who is smart that thinks about this â patents and such. I can do it at some point and I think we will be able to achieve it. DOS being fairly cloned has had a dramatic impact on our pricing for DOS. I wonder if we would have it around 30-40% higher if it wasnt cloned. I bet we would!
Second, we have to make windows really good â not in the ways we have assig to os/2 but in ease of use approaches and software environment richness approaches. We will need smart people to adavnce the product â smaller, faster for the kernel but smarter for all the interface pieces. Apple isnt dont what they should in this area and we can. I am not saying to tilt back towards windows or even to keep particular people but rather to keep what we have and focus it on refinement of speed/size and interface stuff. We may need to do some more software evangelization in certain catergories.
WE have to make windows like DOS is today â basicly new applications all are done for it and people assume it.
FInancially if we havent tied our hands we need to be able to double the oem price or go after a VERY agressive retail strategy. The things we cant afford are:
1. Not being able to improve the product freely. In other words getting into the mess that Os/2 is in. Windows is still not as nice as the mac and it would be easy to be a lot better. Win 3 goes a long ways but we cant get IBMed on this one.
2. A cheap royalty to IBM that means if other oems refuse to per system license we end up getting very little.
3. Transferring a lot of money to IBM.
4. Not positioning windows agressively.
If we hold windows back we wont double our oem revenues. I know os/2 is important and we need it to get UNIX apps ported and to fight the high tech and server battles. But windows is DOS for twice the price and we need it to be pervasive.
I am afraid of a structure with IBM where we are in partnership for the following reasons:
a) server businesses arent enough of a gold mine to be a huge part of our goal. With IBMs cost structure I would like you to explain to me how a server business can even break even.
b) I am hard core about not putting DOS/win into the financial partnership If only IBM users buy their package or that % then a low roaylty is ok but how do we avoid IBM holding us back on windows. Maybe I am paranoid.
c) Lan business is too fast moving to let them write all the code
d) Once os.2 takes over I see our profit per machine actually dropping and never increasing again. This last point I meant to spend more time on. Os/2 will dominate eventually and it should allow us at some point to double again (is that asking too much? maybe) or at least grow per unit more than the UNIX machines cut into the desktop market so we grow with the market.
One question I have steve is:
Shouldnt we let the slow pace of negotiation continue? I think that is in our best interest.
Shouldnt we force them to explain how to make money in the server business since they are the ones who put so much into it?
IBM never knew our plan and if they did they shouldnt like it â our old plan (DOS) which is the current financial success of the companyâ sell cheap to IBM and make money from everyone else. I prefer an OS.2 deal like that a lot more than a joint venture. OF course it makes us sensitive to 2 phenomena I worry a lot about: a) cross selling and b) IBM increases market share. A is the worst since increases in royalties will force some oems to rely on A. I dont know how to solve that. If you feel we have to "joint venture" with IBM I think we give up growth opportunity.
SOmehow this discussion is the most important one for the companies busines future and I guess I should think about it more. In any case there are some thoughts. You can call them greedy. Its funny but I always thought other people in the company would end up pushing me to think more about thi financial upside stuff but it never happens. I was supposed to get to be a product guy with people pushing me to charge enough but when it comes to oses I feel more guilty than anyone that we dont have the clear plan to make 2x as much in oses. Os profit is our past present and future so lets you and I figure out how to double it without IBM problems. No easy job!
[This document is from Comes v. Microsoft (2007).]
Further listening from Acquired: Microsoft Volume 1: The Complete History & Strategy of Microsoft (April 21, 2024)
Previously: Bill Gates: âour strategy for the 90âs is Windowsâ (May 16, 1991)
If you upgrade to a paid subscription, youâll receive access to the full archive of internal tech emails, with 200+ documents from Apple, Google, Meta, Microsoft, OpenAI, Tesla, and more. Youâll also support our work: every year, we track hundreds of court cases and review more than 10,000 filings to bring you @TechEmails.
Wow, Elon sounding reasonable. A different man.