HNHacker News
TopNewBestAskShowJobs

nullsense

1,429 karma · joined September 5, 2020

submissionscomments
nullsense··on ChatGPT and Wolfram Is Insane
The problem I run into on HN all the time is I read exchanges like this and I don't know enough about said topic to verify if either party is right. I guess what I'm trying to say is there are lots of humans who say smart sounding things that to me at least seem believable, but another human disagrees with an equally believable sounding reply and I just go with my gut as to which one I think might actually be the case.

Which human is hallucinating? I don't know. Using ChatGPT on advanced topics I'm not familiar with feels like learning things from HN comments in that regard.

nullsense··on OpenAI CEO 'feels awful' about ChatGPT leaking conversations
If he didn't, we should either not trust him or be very unhappy he's in that job.
nullsense··on OpenAI tech gives Microsoft's Bing a boost in search battle with Google
All this will be a solved problem soon enough.
nullsense··on Bard uses a Hacker News comment as source to say that Bard has shut down
I'm actually finding utility in ChatGPT more and more every day.
nullsense··on Show HN: Recursive LLM Prompts
Then you modify the system prompt to act as a Socratic tutor and you've created a Lead Engineer.
nullsense··on OpenAI to discontinue support for the Codex API
With how fast everything is moving I wouldn't be surprised to see "run a codex like LLM model via torrent" in reaction to this.
nullsense··on Run 100B+ language models at home, BitTorrent‑style
Let's watch the world burn!
nullsense··on Shoshikantetsu
I always joke with my wife that my favorite yojijukugo are 問題解決 and 安全第一 which are literally just "problem solving" and "safety first" haha.

My favorite actual yojijukugo is 点滴穿石 which water dripping on a stone in the same spot slowly wears it away, which just means if you keep doing what you're doing you will eventually make some noticeable progress in the long run.

nullsense··on “Here's the number I used to win the lottery“ –Entrepreneurs giving advice
>7. be lucky

You have to make absolute certain that you remember to do this step and to do it correctly, as well as at the right time, or it won't work.

nullsense··on Facebook LLAMA is being openly distributed via torrents
Reminds me of the time my son's teacher gave a lesson on fire safety telling the kids to not take anything with them and just get yourself out quickly. He realized the implication would be that his entire plushie collection would burn and after that he was inconsolable for the rest of the day.
nullsense··on Facebook LLAMA is being openly distributed via torrents
I made that mistake. One of the most contrived, boring, sappy, fake feeling movies I've ever watched.
nullsense··on Facebook LLAMA is being openly distributed via torrents
It's going to be fun watching the plot of Real Humans literally come to life.
nullsense··on Japan was the future but it's stuck in the past
It does and it doesn't. To them it's a trade-off and they're pretty aware of the upsides and the down. They really like the upsides and they all complain the downsides, but they kind of accept that to have the things they're really proud of as Japanese that in part stems from doing all the things that grind on them.

One of the best examples of "Japan works for Japan, except when it doesn't" is the intersection of mental health and the Japanese way of solving problems. It's important to keep the peace with people, even if you're only pretending to, and that's a higher value than determining what's actually true in a given situation. So, whereas Western people try to get to the bottom of things and figure them out which requires ruffling some feathers, Japanese people very often simply ignore problems in the short term because it keeps the peace while accepting the long term outcome will blow up into a much bigger problem, but oh well because "there's nothing we can do about it". This kind of behavior is super perplexing as Westerner until you grok how it works on a fundamental level as you try to solve problems in a Western way and people wind up getting upset with you, though they often don't say it directly to you. The mental health situation there sucks because they don't know how to help family members who are experiencing difficulty, so the situation goes from bad to worse but of course "there's nothing we can do about it" and "there's no point talking about it" and this is the best thing to do even if it's all just to pretend everything is fine right. There's a reason they have hikikomori phenomenon. And sadly around half of those in mental hospitals could have avoided it if their family simply helped support them properly during their time of difficulty.

Source: watched my sister-in-law who was by all accounts a normal, and amazing human being, legit be driven insane by Japanese society.

nullsense··on DALL-E for Playlists
It's missing the "I feel like crap because I got yelled at for the 1000th time by the same person for doing the same thing wrong" mixed with the anxiety of "I know from experience I can't avoid that happening again, but how I can I avoid that happening again?".

The themes I can detect in the playlist itself are "I don't know quite what this is, but that itself is interesting, so my brain is just going to surf on that for a while", "nothing's going on, so I started daydreaming and my brain is in that comfortable relaxed happy place", and "oh oh oh, Yes yes yes! Mode"

nullsense··on Ask HN: Do you hate software engineering but love programming?
This. The most succinct way I've ever heard it formulated is the quote "not everyone, in the final analysis, actually wants their problem solved"
nullsense··on The 'buy now, pay later' bubble is about to burst
I think the ability to manage money well and not get screwed over by credit facilities is highly correlated with ones level of trait conscientiousness. It's less about skills per se than it is about that particular trait. I would wager those high in trait conscientiousness naturally tend towards frugality and budgeting and those low in trait conscientiousness tend towards impulse buys and poor financial planning.

My wife is 99th percentile for trait conscientiousness and I'm 1st percentile. I'm amazed at the amount of money she's able to save and her level of discipline and she's generally gobsmacked at the amount of money I waste and the random shit I buy.

Having the information about what one should or shouldn't do is only part of the equation. 8t also needs to be something you can apply, and sadly some of us just can't.

nullsense··on OpenAI starts testing ChatGPT premium
Yep. I attempt to educate him on epistemology as much as I can. He's a relatively skeptical kid, but he's also stubborn as hell and he gives somewhat of a recursive answer of "you just Google that too".

I try to impress upon him too that some things take hundreds of hours of explanation before you know enough for it to be functionally useful to you, so you can't always just Google everything. You have to learn things and build your foundations up. Alas he hates the idea of learning anything.

nullsense··on The 'buy now, pay later' bubble is about to burst
>Someone who is financially disciplined is leaving money on the table when not using CCs.

The trick is knowing whether or not you're financially disciplined. And usually one finds out the hard way. It often involves a credit card or credit facility of some nature.

nullsense··on OpenAI starts testing ChatGPT premium
It's called trait conscientiousness.
nullsense··on OpenAI starts testing ChatGPT premium
My son is 7 years and he routinely asks me "why learn anything when you can just ask Google?". I can only imagine the impact this will have on that attitude over time.
nullsense··on Your tech stack is not the product
We've never met and we probably never will, and it's a virtual certainty I'll never see the code of which you speak. But know this... I appreciate it. And I can imagine it. I certainly empathize with your plight. The feeling of a job well done is all the validation you need.
nullsense··on Ask HN: Strategies for working with engineers that are too smart?
I think one of the key tensions in this kind of scenario is individuals combination of traits massively influence the way they view the code and they're generally blind as to what the world looks like to engineers with a different combination of traits.

Just because someone has a higher IQ doesn't make them objectively better on all measures. It usually does mean they have cognitive access to things that others do not, and this can be beneficial as well as detrimental. It's seriously beneficial when they are able to come up with a simpler, more elegant solution when the rest of the team can't where that solution ratchets down the level of complexity and/or the cognitive load on the team. It's seriously detrimental when they propose solutions that do the opposite simply because they can conceive of it and it solves some problem they think exists and/or that needs solving and ratchets up complexity and increases cognitive load.

I like the example of the engineer arguing for less whitespace because he personally doesn't need it where others do, but he's unable to see the effects it would have on other people were it to be removed and how in aggregate that would likely result in a worse outcome.

There's an engineer that I work with occasionally who is like this. He loves complexity and absolutely thrives on it. It's like oxygen to him. He's very smart. He's also a 20+ year veteran, 10+ of which is with the current company. There are things which I know he is uniquely capable of producing, and in such a scenario I want what he would produce as it would be significantly better than what almost anyone else would do. However, 99 times out of 100 I want what almost anyone else would produce than what he would, as he optimizes only for himself and not the wider team. We have an upcoming overhaul of a legacy system that were both chomping at the bit to get stuck into, and I've been told to "get in line", but I'm not just going to lie down and take an attitude like that when I know what the result might be. Needless to say I think it will require being both firm and diplomatic.

nullsense··on Microservices are hard
It turns out the One Weird Trick is just considering the context.
nullsense··on Theory-building and why employee churn is lethal to software companies
Yeah I've done this too.
nullsense··on Theory-building and why employee churn is lethal to software companies
Really don't wish to go back to this. We don't have QA teams or QAs where I work anymore. We can get changes to prod much faster and our test suites are growing to be pretty high quality for the stuff our team owns.
nullsense··on ‘That girl is going to get herself killed’ (2022)
That was wild.
nullsense··on ‘That girl is going to get herself killed’ (2022)
>Random story but I just wanted to share that you can prepare all you want and sometimes life still tries to kick you in the butt.

Probably the most useful mantra from the prepping community that gets burned into your brain is "two is one, and one is none".

nullsense··on Tell HN: Airbnb has no system in place when a host cancels on a renter
Up to 4 days probably accounts for long weekends.
nullsense··on How to store your app's entire state in the url
This has to be my favorite version I've ever heard.
nullsense··on I don't read web articles anymore, but I read books
I tried blinkist but it never stuck. I think long-form stuff is better for developing semantic memory. Sure you can get to the point quicker with shorter form content, but tends to be less sticky.

I haven't gone back to blinkist to try this but I think a way to build up semantic memory with blinkist would be listening to every single title in a specific category to go really deep on that one category. At that point you're doing the equivalent of reading a full length non-fiction book anyway, so you haven't saved time, but at least you could triangulate what the core ideas / themes are through the entire category and really get the lay of the land whilst also getting the core messages deeply ingrained.

← PreviousPage 7 of 29Next →