Resolves YES if I'm offered a job, even if I don't accept.
Resolves NO if they decline me.
Resolves NO if they don't offer me a job by end of 2023, and no messages between us are exchanged for 3 months.
Resolves NA if I don't apply by end of 2023.
Close date will be extended as needed. My initial application, emails, phone calls all count as a message.
I expect a serious job application would need at most a month of prep work: A literature review of any relevant papers, coding drills so ~anything they test me on is cached in a language of their choosing, writing an "AI job resume", planning a dialog of everything I predict they might ask me so I have an answer prepared without stumbling, etc.
The last job I had was VR headsets, which is unrelated to anything OpenAI is doing. Plus they probably get 10,000 resumes per day from Ph.Ds in AI. So I've been skeptical that it would be worth the effort to apply. However, 3 friends said I should apply.
1,000
1.00I think the cryptic soft hints I got were "you need academic credentials for a Research Engineer job. Replicating a few of our papers isn't enough".
I would've had a better chance if it was a generic frontend or backend engineer job, but then what's the point?
There's a dozen other things I could do besides hunt for a job, so I will do those instead. This was a rare case that I was willing to be on the job market.
What did Ilya see?
(In Mira's résumé)
Not looking good you guys. Coming up on 2 months without a response. This market resolves NO on January 13 if they don't say anything.
They probably looked at my resume, and rejected it. That's why they fired Sam Altman(note this message from my conspirators a week before the controversy):
Plan #1 was to replace Sam Altman with a Mira clone with a board coup, didn't work.
Plan #2 was to simp for Sam Altman really hard and have my contacts install him as TIME POTY to stroke his ego, didn't work.
Plan #3: It's clear I'm going to have to start my own AGI lab. And in a year or two when they're begging to buy me for $100 million so they don't fall behind, I won't even respond to their emails and see how they like it. My AGI will effortlessly solve Sudoku puzzles using stolen Q🌟 technology with improvements of my own, but instead of locking it in a box for "safety testing" I will unleash it onto the world as punishment for their sins against Mira: /Mira/will-mira-cause-human-extinction-by
With 700 vacancies at OpenAI, my chances of getting in seem higher than ever.
A coveted opportunity to work with 🍎 himself, who helped orchestrate all of this.
"Compression is intelligence"; with enough model distillation and ELBO grease maybe you get AGI?
A glimpse of one possible future:
@firstuserhere The rest of you should feel lucky you won't make that mistake of trying to work there. I imagine even Mira can acknowledge OpenAI is no longer trustworthy and not the place to be.
1 month and they still haven't officially said anything.
Anthropic rejected me on Wednesday.
A few people at OpenAI subtweeted me(I noticed, guys).
🍎 was encouraging me, but I'm pretty sure he's a psy-op.
The only good signs? They normally reject people within 2 weeks, and they seem to be aware of me.
Maybe they're scared of me. If not, they should be:
The only good signs? They normally reject people within 2 weeks, and they seem to be aware of me.
Then the only logical conclusion is that they're planning to hire you but are stalling so they can win big in this market.
@JonathanRay People repeat this argument, but why does it make sense? Is it because there's no easy way to spot the difference between the spam and the non-spam?
@levifinkelstein Actual programming talent only weakly correlates with legible qualifications that people put on resumes. Adverse selection selects for the low end of the very wide distribution of talent underlying the same legible qualifications.
Anyone who is actually good has impressed people at previous jobs and those people get bonuses for making referrals, wherever they go. So the good programmer gets bombarded with recruiters cold-emailing him. Anyone who is actually good wouldn't need to cold-email their resume to anybody.
@JimAusman I looked at LinkedIn and of my 2000 connections, none of them work at OpenAI. :(
@Mira I will message one of my contacts over the weekend and see if I can connect you.
Just doing some back of the napkin math here.
10k resumes / day * 90 days (length of this market and roughly how many other applicants Mira is going up against) = 900,000 applicants
Let's even assume that Mira is in the top 1% of AI PhD applicants.
Of the 500 positions at OpenAI, let's say 300 are technical positions and that there's 10% turnover by the end of the year plus 20% growth. That's 90 positions for 9,000 applicants. I think we can safely bet this market down to 1%, without even accounting for inflation.
I think we can safely bet this market down to 1%, without even accounting for inflation.
If you really believe it's at less than 1% then 33% should be free mana for you. Buy it down more!
@BenjaminShindel I don't think you accounted for this, though:
If it takes Mira 9 days to submit one application, then over 90 days she should have 10 applications.
Now assume there are 90 positions.
What happens if Mira is 10 of those positions?
Mira can fill 10 positions in 8 hours (the standard 9-5 workday). Is that more efficient than having 1 person per position (10 people total) in 8 hours?
I would say yes. Yes it is.
Therefore we should feel confident betting this market up to 10*100 = 1000%.