your m1 is a beast of a machine, especially for running pre-trained models. you should be able to run any of the image identification models no problem, and in real time.
what's even more mind boggling is that library is more or less publicly accessible _now_ via a home internet connection or a connection from your... friendly neighborhood library! but you're right, having unmitigated access to that is even more incredible.
yes. on modern hardware, OBS is relatively lightweight until you add several video sources. for coding you should be fine if your laptop is less than 5+ years old
this phenomenon also happens within businesses and in a more bizarre fashion. look at all the layers of management a startup will slap on at some point in it’s life because it’s commonplace at other companies and Google. look at the processes like Jira, stand ups, sprint planning, etc etc that the management class will institute and defend because it creates work and meaning for their position.
i think your imagination is extremely limited. we have essentially no understanding of the brain or consciousness at this time. making strong statements about the limitations of machines is just naïve
if you decide to experience the entire world, all the time, there's no respite from the awful. the human mind just isn't capable of handling that well.
1: yes. you get cash for your shares.
2: the board is representative of the share holders, like your US Congress person. they have a fiduciary responsibility to give the share holders value. the only reasonable way they could blow up this deal is if twitter had an incredible roadmap with a very good path to matching the value or exceeding the value of Elon's offer. no such roadmap exists, apparently.
yes and no. the last company I worked at is a feature factory. They have senior, and staff engineers, on various product teams that have 0 technical influence. What they expected is you to somehow increase quality of the product via code review and writing design documents... but the quality of the product was being destroyed by a handful of people making all of the technical and architectural decisions and forcing their platform on the rest of the org. Upper management doesn't have the technical expertise to see this and thinks this approach will lead to some sort of unified performance and scalability. Maybe it will down the road, but if I can't change out the underlying database, ORM, or web framework for something that is just _better_, I'm ultimately just working on the assembly line at the feature factory as a Sr/Staff engineer.
i kind of like the idea of taking them through your codebase, though. allowing the interviewer to concoct problems leads to something i'm very passionate about, and that's attempting to stifle the ego involved in the interview process. for whatever reason, intellectual prowess, and displaying it, exposes itself in interviews. the interviewer will ultimately end up trying to dominate you intellectually to satisfy their own ego. it's pretty pathetic, and rampant in whiteboarding.
bizarre tech decisions have led to insane costs with 0 customers. launch is about 2 years behind. having a lot of trouble hiring senior engineers, when they do they don't listen to them. they have a strange platform that every product team must use. tech leadership has never launched a successful startup before. they have over 300 employees, maybe 400 now. 4 levels of management already, and a _lot_ of them. it's independently funded for now, and has the potential to keep on going like this as long as Dave D. doesn't pull the plug.
I’m mostly familiar with STEM PhD’s, and do not have one. However, my understanding is coursework is over once the PhD starts and you’ve earned a masters. You’re in the lab doing research on your thesis, and often teaching classes.
haha — my email is in my profile page if you really want to talk about it. i have interjected at previous companies where it was possible to influence the process a little bit. my biggest question would be “how could you do it differently than triplebyte when they first start?” they started by interviewing people themselves which is hard to scale.
right — it’s certainly not FAANG specific but seems to arise from these styles of interviews. maybe it’s lack of interviewing training or standards? i think the standard now is, “pick your favorite hard problem and have a candidate do it on the whiteboard.” which presents this opportunity for an interview lacking compassion.
yup — I’m a bit old school and keep trying to do this. It’s very “hard,” you can face rejection, but it’s way more interesting than living in a bubble of your work place.
the opportunity this presents is getting lunch with people that aren't ephemeral and transactional. you can experience something OTHER than work on your lunch hour, expand your horizons, and build meaningful relationships. of course you will on occasion get one of those relationships from the office, but they're the exception to the rule imo.
this is the in-depth reply that illustrates my flippant comment. there can be an inherent competitiveness in white-boarding interviews. the interviewer can implicitly want to show their superiority over the candidate. i think that's the core reason behind a lot of overly difficult problems being offered that have no analogous presence in the day to day work. how many times have you written a graph traversal algorithm on the job? or implemented a geofence in 45 minutes, from scratch, without a search engine?
i whole heartedly agree on the pair programming approach being practical and yielding good results. i think you can skip exposing the candidate to the internal codebase, and replicate an internal problem in a more generic and high level way.
I think we're on the same page. It's a matter of $$ for speed in the sequencing and analysis phases. However, the big problem that Isaac Asimov posed and is still relevant, is the understanding of the results. Having your entire genome sequence is achievable now, but understanding it isn't. As far as I know 23andMe and other services only look for a subset of well understood variants and that's it.
how much do you want to spend? you could do it in a couple days if you use a few sequencers. how long would it take to assemble? how much do you want to spend, etc.
how useful is it? not very at the moment. most analysis is still around specific variants that are understood. most function is not understood.
of course — so you have to make it accessible to the candidate. using the internal code base is probably not the best choice. we had them write up a small API with specific requirements in the tech of their choice, and tried to have someone programming with them familiar with their choice of tech.
making the interview process similar to the job is a good first step. when I worked at Invitae this was my proudest achievement from my time there. I suggested that we change the interview process to a paired programming session that were situations from our day to day work. the candidate brought in their own laptop or was provided one (their choice). the session was 90-120 minutes and didn’t involve leetcode hard problems, just practical day to day stuff a web developer might run into on the job.