1,043 karma · joined July 6, 2018
As someone who does very highly value this way of thinking though, how would you try to detect it in an interview setting as an interviewee? That’s like asking them if they have good culture.
None of the open models available to these local tools benchmark as well as the closed OpenAI model but the benchmarks are weird. In my own poking around, some models are definitely less coherent or more detailed for some specific prompts but it’s mostly the same.
I encourage this hacking spirit in them though and let them have a whole day if they find a new workaround. Right now, I have the unlock code set to all-zero but they haven’t figured that one out yet!
Best advice for getting them off devices I have is to take them away. If you need them distracted for a bit, a larger screen like a real TV or desktop doesn’t seem to hold them as tightly too.
Trying to be more practical and a touch less sarcastic for a minute… I find, on average, I tend to get about two hours more work done on days where I spend at least 45 min in the gym and that includes the hour or more loss from the activity itself. Skipping caffeine on the other hand, doesn’t have any effect unless I’ve missed sleep the night before.
If you are thinking of dipping your toes into supplements though and don’t really want to exercise or sleep, I’ve found Creatine to be really helpful in that state. Definitely a stronger effect than caffeine but takes a lot longer to kick in!
TLDR: Your starting point is exercise, sleep and sunlight. Nothing after that really has as large of an effect.
If you’re thinking of the stress from that as a kind of training regiment, what would someone look like who thrives in that environment? What does success look like? Less coddled is an interesting take on that.
If this collection is trying to solve the problem of having large tagged and sorted data for training, even if it can’t be commercialized, it might also be introducing another where the data still needs to be weighted and filtered. Not to knock on Romance specifically but even training on all scientific papers ever written, whether peer-reviewed, highly cited, retracted, debunked, etc. might explain some hallucinations it ends up making.
I think we’re more likely to advance on the production side with improving yields and reducing waste rather than trying to reduce consumption though. This metric wouldn’t really show progress in that direction.
Talent is transferable skills. Getting good can’t come without practice but you don’t have to be practicing the exact thing.
Maybe the hot take here is that the best way to ruin your own business is to automate credit cards?
Is there more to the story here that I’ve missed?
As far as I know, all of the sugar alternatives have some (usually neuro?) downside that makes them scarier to me than sugar!