OpenAI is Hiring. This Time, Not for AI and ML
Gauging the fullest ambition of AI-powered tech empire
I am not an OpenAI recruiter — just a curious follower of tech companies.
Through sheer curiosity, I was browsing OpenAI’s website. I came across a link that led me to an OpenAI Residency page.
It’s a 6 months program aimed at attracting researchers — across fields such as physics, mathematics, and neuroscience.
The new-age AI unicorn is no longer just a tech company.
About OpenAI Residency:
The position is located in San Fransisco. It spans 6 months. However, during the tenure of residency, OpenAI will pay a salary it claims as competitive: $2,10,000 / annum.
Converted to a half-year commitment, it comes to $4,20,000 / annum. It’s a decent figure for sure.
Drawing from that, OpenAI stresses that it’s not an internship. This means that researchers are there to give something valuable, not just be trained and be done with the hottest tech company of the 21st century.
Why Other Domains?
Firstly, GPT is not just a machine learning product. It’s an NLP product, made by expert linguists who tutored data scientists how to tame language.
In its current form, even with its flying color exam grades, general purpose GPT is only good as a chatbot.
One or the other way, any LLM would need domain expertise. Statisticians, data scientists, ML engineers — all such specialists can’t provide it.
What’s in it for OpenAI?
Since it went closed-source, much has been under the wraps at OpenAI. With Microsoft being its largest investor and benefactor, the possibilities are endless.
Microsoft is a dominant name in the government and enterprise software domain. With GenAI at its side, there is so much it can achieve, beyond selling solutions that are married with Azure. Many large consumers hate Microsoft for its ugly packaging, and the case of GPT won’t be different.
Microsoft understands this. AI in itself is nothing unless it is rooted + trained in industry/domain-specific knowledge.
Presently, investment in the AI sector is making it look like there is no tomorrow. But naturally, most of this is driven by hype. Most budding AI companies are either quite vague about the execution or highly dependent upon GPU-guzzling API-based AI usage.
To utilize GenAI both cost-effectively and meaningfully, AI consumers need to work around their pre-existing knowledge bases. For smaller companies, this is almost meaningless — unless their offering relies on public datasets without privacy and licensing issues. For bigger firms, this could necessitate restructuring the knowledge bases for consumption by AI teams.
This wave is bound to break, and Microsoft would obviously want to save itself from the Dot-com bust 2.0. It knows how to prepare the next tier of sellable AI product suits.
OpenAI’s growth trajectory is far from optimistic if they can’t offer a USP to their huge, non-tech corporate players.
OpenAI is a novel challenger. But there is already a giant wrestler in the room.
Google:
While OpenAI rode on sensationalism, the Bard owner hasn’t been sleeping.
To say “Google’s Palm-2 is revolutionary” is an understatement. It is a worthy challenger to GPT-4 in several areas such as puzzle-solving, bug-spotting, and translation. Med-PaLM is its version designed for the medical domain.
Healthcare is a trillion-dollar opportunity, and Google already has a headstart: Its subsidiary DeepMind has a product called AlphaFold which can predict protein folding. A brief background: Protein structure prediction is one of the most important goals pursued by computational biology. Its direct outcomes are improved, cost-effective drug design and the design of novel enzymes.
That explains why OpenAI is interested in biologists. In fact, in its call to neuroscientists, there is a visible urge to play along with Elon Musk’s Neuralink and its friends. The possibilities here are endless. The partnership could bear a novel, AI-powered Neuralink app store — who knows?
That leaves Physics. Why OpenAI is interested in physicists? Well, the reason is again Google’s acquired AI child. If you remember, DeepMind is the company that became famous via AlphaGo — a computer system that defeated Go world champions.
DeepMind might have plans to create AI-based bot gamers that follow the laws of physics and game mechanics. Another much more lucrative possibility is to create realistic motion pictures without the need for designers. If that fructifies, Google will happily want to fight Netflix and Disney in their own arena, and maybe even beat them.
OpenAI wants its own player in the ring, with the full backing of its rich benefactor.
What’s in it for researchers?
A career break from the hottest tech sector company — that too in the heart of Silicon Valley. Anyone would say it is unbeatable.
But the question is: Do they see a path past that? Lining up for OpenAI, they would be definitely expected to write Python code at some point, though the opportunity doesn’t state it.
If not today, they might walk that path with some other employer tomorrow.
I am not saying scientists can’t code. Or they won’t be inclined to. It’s a question less of their adaptability and more of a mindset. More so, if they are just beginning their careers with OpenAI.
How will the experience shape their own perception of their impact-making capability?
For them, the real question to answer is: How many companies are already in the process of creating a talent pool for cross-domain professionals like them? And are they all really worth their salt?
Conclusion:
Is OpenAI’s fellowship for academics really worth it?
I am perhaps the worst person to answer this. I have never been in academia, let alone research.
With academia bleeding across the globe, it’s a buyer’s market for corporates.
When GPT-2 arrived, I predicted a far worse turn of fate for writers. The world just witnessed a 148-day-long strike by Hollywood writers. The writers won, and I am happy for all of us. But there is no reason they won’t be enslaved en mass.
Big corporations will find a way to get rid of the workforce. Because they have the software by their side.
It’s a huge, closed-source monster, whose head and tail are invisible. All we can see and touch is his huge stomach.
And with that, it is eating the world.
Want to get an email every time Pen Magnet publishes? Click here to join his subscriber list.
Pen Magnet is the author of the popular senior developer interview eBook:
Comprehensive Approach to Senior Developer Interview (40+ example questions)
