It would be my life's dream to spend 80 hours per week coding without having to communicate with others... but no one is an island...
You really have to have a passion for coding to put in the hours and be very good at it. Incredibly rare, believe it or not. Lots of people think they are good coders but this is another level. Proof is in your commit/code review count/async comms being 10x-100x of everyone else in your org, and it's clear you're single-handedly enabling delivery of major projects earlier than anyone else could. Think of the pressure of doing this continuously.
You have to watch out with that.. I've seen whole projects pushed through by management where no one else was involved enough to review normally, but everyone had an interaction that implied they had only seen the top of the iceberg of problems with it.
That's a nice insight. I have been in that place many times, I was overfitting on my own imagination.
When it comes to sports it's fairly obvious what outliers look like and well accepted that they exist. I don't see a single reason to believe, that the same would not be true in every other walk of life or thinking that OpenAI just got lucky (considering how many people are trying to get lucky right now with less success in this space).
There are extraordinarily effective people in this world, and they are sparse and it's probably not you or me (but that's completely fine with me, I am happy to stretch myself to the best of my abilities).
In my company, 80% coding for a senior SWE is rare. But if they deliver, management will give them some slack on the other evaluation axis. I have colleagues who work almost by themselves on new high impact projects. This has many benefits. No need to argue about designs, code reviews (people just approve blindly their code). The downside is that you need to deliver.
I don't know this guy in particular so I have no clue though.
What also happens is regular developers (like me) want the same treatment as if they could end-to-end deliver "if they only let me", but many times can't, and actually need the structure and processes of a team. I've seen this freedom not working at all.
- an equalizer (entire team treated the same)
- a confidence booster (approval of others gives feeling of having done well)
- a way of distributing information (everyone is aware of all other team work)
You can run a team as a form of "competitive sport", and race everyone against each other; who churns out most "wins", and helpfulness, non-code-work, cross-team work are "distractors" to that objective hence undesirable and definitely not rewarded.
If the personalities in your team are "right" then this can work and by striving to best each other, all achieve highly. Have a single non-competitive person in there though... and it'll grate. Forcing a collaborative element into the work (whether by approval/review procedures, or by things like mentoring/coaching, or even just to force briefings to the team on project completion) creates a balance between the "lone crusaders" and the "power of the masses". Make the loners aware of, and contribute to, the concept of "team success", and give the "masses" insight into contributing factors of high individual performance.
The topic at hand is “how did a high level engineer got to focus on programming”. And I am saying that the reason has to do more with his influence and role within the organization, rather than other reason.
For a certain definition of "software": when only doing one training run costs an 8 digits sum (requiring hardware one order of magnitude more expensive than that to run) I kinda dispute the "all they do is software".
It's definitely not "all software": a big part of their advantage compared to actually free and open models is the insane hardware they have access to.
The free and open LLMs are doing very well compared to OpenAI once you take into account that the cost to train them is 1/100th to 1/1000th what it costs to train the OpenAI models.
This can be seen with StableDiffusion: once the money is poured in training the models and then the model made free, suddenly the edge of proprietary solutions is tiny (if it even exists at all).
I'd like to see the actually open and free models trained on the hardware used to train OpenAI: then we'd see how much of a "software edge" OpenAI has.
And my guess is it'd be way less impressive than you make it out to be.
It would seem like you're talking about what "software edge" OpenAI has in the future, when others have caught up, while parent is talking about the existing "software edge" OpenAI has today, which you seem to implicitly agree with, as you're talking about OpenAI maybe not having any edge in the future.
Because most 10x engineers recognized by management as such are characterized chiefly by building out shoddy software extremely quickly that only they can understand.
In a similar dynamic, Doctors that are scored highly by patients often have pretty bad medical outcomes.
I asked him whether as a boy he had speculated much about his gift. Had he asked himself why he had this special power? Why he was so bright?
Dyson is almost infallibly a modest and self-effacing man, but tonight his eyes were blank with fatigue, and his answer was uncharacteristic.
“That’s not how the question phrases itself,” he said. “The question is: why is everyone else so stupid?”
I do not wanna be flippant here: Obviously having easy access to money and a good standing with the right people is making things A LOT simpler, but other people could have reasonably convinced someone to give them money to built the same software. That's what VCs do, after all.
Regarding the rest: Feels very much like a different topic. I'll pass.
Management and leadership of a team has a way bigger impact than any single individual contributor could ever have. Humans are generally limited not by intelligence but by motivation and vision. Directing people to achieve what you want is what allows the scaling of innovation.
Hero worship is a very human thing, but unscientific.
1) Moves fast, flexes their authority to sweep small stuff under the rug until it is out of scope and can be "fixed real quick" later. Often leverages many subject matter experts through effective and persistent communication and learns quick enough to get PRs through the door (that sometimes need "quick" fixes later). Enjoys selecting items that benefit their career the most, at the expense of others on their team. Mentors only enough to onboard and increase his team's yield, not to aid their careers. Fueled by the recognition and validation of peers through PR/project completion.
2) Gets shit done, is the SMI themself. Solo code cannon, but PRs go in clean, beautiful to look at. May not get along well with some but not necessarily abrasive to work with especially being part of their direct team. Can be a great altruistic mentor if they spare 5% of their time. Enjoys what they do, and the technologies they work with. Fueled by personal satisfaction in their achievements, and in uplifting their team.
https://www.cnbc.com/2017/04/21/why-apple-co-founder-steve-w...
My take: He’s the Keanu Reeves of tech (or Keanu is the Woz of the film industry). The world can use more of this.