sankalp's blog

In undergrad, Shoulto would study papers regularly from 10pm - 2am and on weekends would spend 6-8 hours reading papers -> Insane work ethic

On Research and Taste

Below are the **verbatim remarks the speakers make about (a) agency and (b) *research taste***, grouped and lightly trimmed for readability. Ellipses (…) show places where a few words were removed, but the order and wording are unchanged.

1 Agency — driving work end‑to‑end

# Exact quote
A‑1 “I just don’t get blocked very often… if I’m trying to write some code and something isn’t working… I’ll often just go in and fix that thing or at least hack it together to be able to get results.”
A‑2 “That’s arguably the most important quality in almost anything — pursuing it to the end of the Earth and whatever you need to do to make it happen, you’ll make it happen. If you do everything, you win.”
A‑3 “From my side… agency in the work [has been crucial].”
A‑4 “I’ve been very good at picking extremely high‑leverage problems… then going ‘okay, well I’m just going to vertically solve the entire thing.’”
A‑5 “I was hired… as an experiment in trying to take someone with extremely high enthusiasm and agency and pairing them with some of the best engineers that he knew.”
A‑6 “For the people who assume the job board is totally mechanical — this is not how it works… people are looking for the kind of person who’s agentic and putting stuff out there.”
A‑7 “Specifically what people are looking for there is two things: one is agency… and the second is the ability to do world‑class something.”
A‑8 “There are no adults in the room… you have to decide what you want your life to look like and execute on it.”
A‑9 “One final thing… we talked a lot about agency, but one of the most important things is just caring an unbelievable amount… going and fixing things that aren’t your responsibility because overall it makes the stack better.”

2 Research taste — choosing interesting & fruitful questions

# Exact quote
T‑1 “…during my self‑experimentation I was obsessively reading papers every night… across NLP, computer vision, robotics… you see all these patterns start to emerge across sub‑fields.”
T‑2 “My research agenda and the interpretability team seemed to just be running in parallel, in‑line with just research taste, and so it made a lot of sense for me to work with the team.”
T‑3 (On Andy Jones’s low‑budget scaling‑laws paper) “It demonstrated incredible engineering skill and incredible understanding of the most topical problem of the time… as soon as that paper came out several labs wanted to hire him.”
T‑4 “It’s amazing how quickly you can become world‑class at something because most people aren’t trying that hard… if you just go ham then you can get really far pretty fast.”

These passages capture how the speakers frame agency (unblocking yourself, owning the whole stack, manufacturing luck) and taste (cultivating wide foundational knowledge, spotting high‑leverage or topical problems, and caring about elegant questions).

Key points with verbatim supporting quotes

Key idea Exact quote(s) from the transcript
1  Timing & luck enable rapid growth “luck obviously and I I feel like I've been very lucky in …the timing of different progressions has been just like really good in terms of advancing to the next level of growth”
2  Relentless execution & fast feedback loops “we just needed to really execute on them and have like quick feedback loops and like do careful experimentation”
3  Agency: never get blocked, fix the stack yourself “I just don't get blocked very often… if I'm trying to write some code and…something isn't working…I'll often just go in and fix that thing or at least hack it together to be able to get results”
4  Pick extremely high‑leverage problems “I've been very good at picking extremely high leverage problems, so problems that haven't been particularly well solved so far”
5  Hiring via enthusiasm + mentorship experiment “I was hired…as an experiment in trying to take someone with extremely high enthusiasm and agency and pairing them with some of the best engineers that he knew”
6  Broad reading before specialization creates cross‑field insight “I was obsessively reading papers every night… across NLP, CV, robotics…you see all these patterns start to emerge”
7  Manufacturing luck: more shots on goal “Choosing to go to conferences itself is like putting yourself in a position where luck is more likely to happen…my own way of trying to manufacture luck”
8  What orgs actually want: agency + a world‑class spike “people are looking for the…person who's agentic and putting stuff out there” + “the ability to do world‑class something”
9  “The system is not your friend” → own your path “there's this line the system is not your friend… it's not actively against you... it's just not looking out for you”
10  Caring an unbelievable amount drives quality “one of the most important things is just caring an unbelievable amount…going and fixing things that aren't your responsibility”
11  World‑class is closer than you think if you go ham “it's amazing how quickly you can become world class at something just because most people aren't trying that hard”

These direct excerpts illustrate how the speakers credit their success to timing, obsessive execution, choosing leverage, personal agency, deliberate luck‑making, deep care, and sustained intensity—not merely baseline technical skill.