i don't think i'm the best writer there is, but i do think that, to a certain extent, i find writing easier than most people. a couple of people have asked me how to write, so in a few words, i want to share my answer.

i strongly believe that you can't know what good or bad is until you've experienced a variety of options. when you experience enough of them, your brain subconsciously begins to recognize what you like and what you don't. it does a form of classification. "this is good", "this style sucks" typa classification.

for tech nerds, this is somewhat similar to unsupervised learning in machine learning: it's learning by being exposed to a broad range of unlabeled data. a good number of ai models are trained like this.

to become a good writer, you can replicate this by reading a lot.

when you read widely, you're exposed to different styles, voices, structures, and ways of expressing ideas. without even realizing it or making an effort to, you start to develop an intuition for what feels good and what doesn't. so as you read more books, essays or articles, you add a little more to that intuition.

and you don't have to limit yourself to any particular type of book. i strongly advise people to read as much as possible. read as widely as possible. classic literature, poems, fictional stories. cast your reading net as wide as possible.

ai models use enormous amounts of data to learn how to perform tasks well. so in the same way, you need a lot of exposure to good writing to train yourself to become a good writer.

it would be dumb to say "as long as you read a lot, you automatically become a good writer."

once an ai model has gone through its initial training, it goes through post-training, where its capabilities are tested and improved using feedback and additional data. then, they add reinforcement learning to take this further: the model learns by attempting a task, seeing where it fails, and adjusting.

you have to do the same thing.

so write. write more. get feedback. figure out where you're weak. write again. over time, you get better and better.

no one starts off as a great writer. after months and years of reading, writing, receiving feedback, and improving, you eventually become one. i mean the grading for writing is subjective but you get what i mean…

so yeah, the best way to become a good writer is to read more…

who would have thought?

unconventional thinker

it's not surprising that one of the most common use cases of ai recently has been the automation of job applications. job applications are very repetititive and most people hate it by default. so there has been a great turn up of startups whose entire product is based around "apply to jobs with ai". and with these tools, you can apply to hundreds of jobs in one day.

i think automation of tasks is a great thing and any unproductive task should be automated away. but i will push back on applying to jobs with ai. more recently, people have been complaining about how cold applying is now almost dead because thousands of people apply to jobs and recruiters just can't keep up. a great deal of the applicants for these jobs aren't even qualified. some aren't even interested in the role or the company. but because they use ai tools that have automated the applications for them, they just spam away applications. so if anything, automating job applications has almost killed the usefulness of job applications.

due to a high inbox of applications, recruiters are forced to use ai tools to screen. this makes the job easier for them but i think it also cuts away a lot of people who might have stood a chance if their application was reviewed in context by a human.

i believe job application should be a process with as little as possible of ai interference. the job being recruited for will be done by a human. so a human should apply, fill in the forms and submit their resume after going through the job description themselves to see if the job suits their unique skills and if they love the company enough to want to work there. and of course, a human(recruiter) should read and evaluate the applications as well.

applying to jobs by yourself means you will probably apply to less jobs but instead of asking your ai to spam more, use that time to concentrate on those few ones you applied to. try to reach out to a recruiter or someone who works at the company. research more on the company and look for solutions to any problem they might be facing. for tech folks, maybe use their product to build smth and post on x. i did this for the co-founder of an ai lab and he sent me instantly to their director of ml. i haven't gotten the job yet but that shows the potential of putting more effort on one company.

by the way, this is harder. you'll spend more time on one company than you would applying to hundreds with the ai tools. however, it gives you a better chance at that company than at the thousands your ai bot has applied to.

quality over quantity always!

unconventional thinker