I have already seen the LLM hallucinate prompts from me - in this case, hallucinating being asked to switch to a different programming language - because it wasn't able to complete the task asked for in a satisfactory way instead of giving up and telling me it's not able to do it.
If it doesn't already, I suspect training needs to include those no-solution scenarios and reward not overstepping bounds, or else we're going to see a lot more harmful side effects.
This sounds like fun, but I'm not sure how changing build systems fixes the vendor problem. They will just ship their proprietary blobs built on top of this new build system instead of the old one, and will not change other behavior. Buying from vendors who upstream or minimally open source their work might help.
I'm a little doubtful on the post's assertion that yocto is only needed for "deeply embedded regulated products." For products that have to follow the https://en.wikipedia.org/wiki/Cyber_Resilience_Act coming up, they will almost certainly need:
And I'll echo what some other folks have said here: if you don't need features like that, bundle your custom bootloaders and kernel with a binary distribution like Debian and call it a day.
There's a relatively easy technological solution to mold: buy a dehumidifier.
We bought two after moving to Ireland. Both have drainage hoses. One has a pump and empties into the kitchen sink, the other has no pump and the drainage hose empties into the shower. No more mold problems.
Thank you for sharing. Do your dehumidifiers run all the time? Do they have some kind of auto on / auto off feature? Do you run them on some schedule? As necessary? I am thinking it would be very easy for me to forget unless it is a set it and forget it kind of thing.
Ours has been on constantly for nearly a year. Any decent one has humidity set points - we set ours for 55%. It's a bang-bang controller with a 5% range - it'll run until humidity drops to 52%, then turn off until it rises to 57%. During the winter our single one struggles to get much below 60%, we might add a second next year.
An unexpected benefit (for us, as its not something we're used to) is its virtually "free", as we don't have to run the electric clothes dryer anymore (nearly €4/load). We just hang the clothes on a rack by the dehumidifier and it dries them out in a few hours. My wife is starting to prefer it as its not destroying clothes nearly as quickly.
> An unexpected benefit (for us, as its not something we're used to) is its virtually "free", as we don't have to run the electric clothes dryer anymore (nearly €4/load). We just hang the clothes on a rack by the dehumidifier and it dries them out fairly quickly.
It’s not free as the dehumidifier has to do more work. If you have a modern heat-pump clothes dryer you might be using more electricity by abusing your humidifier like this.
"Free" as in our electric bill was basically unchanged before and after it; and we still get 10-15% lower humidity on average which has taken care of our (minor) mold problems.
I used to be in a house that needed a dehumidifier. I bought one with a built-in pump to empty the water tank, which can be tapped into an AC drain line, sink drain, or directly outdoors, depending on your situation. You can set many of them to run constantly or when humidity is above a certain level.
If you live in a hot/humid location your HVAC is much better at dehumidification than a dehumidifier, since a dehumidifier outputs a ton of heat to operate. But in a cool/humid location, you can definitely rig up one to be decently set and forget. You still need to perform maintenance (cleaning the grill, adding a bit of white vinegar to the water tank to kill growth buildup, etc).
I actually put mine into the HVAC cabinet and used a smart plug and smart thermostat to turn on the dehumidifier and the HVAC fan to dehumidify the house. But that’s way more than you need to do for a small apartment.
Hah, if you ever used the N800 media player you will have been exposed to a tiny bit of my code. Some of the UI polishes and usability tweaks were mine. (Well, someone else had figured out they needed to be done and the bug landed on my lap...)
"/* Here be dragons */" in a particularly hideous d-pointer punching chain must have been a surprise for whomever eventually picked it up.
It's great marketing! But yes. He considers it an error to specify "Use the same knife for the jelly" even though it's considered correct to state "Wipe the knife clean before using it for jelly". The latter statement implies the former, and if you follow all the instructions both are not wrong.
I also consider some of the instructions to be under specified. For example, a piece of bread could be said to have 6 sides, but only 2 of those are helpful for making a sandwich.
I'm just annoyed I lost points for specifying both "Open the peanut butter jar" and "unscrew the lid of the peanut butter jar". The first one's context! The second "more precise" version doesn't specify what you should do with the now-unscrewed lid and the obvious solution, for a robot that takes things literally, is to leave it sitting on top of the jar.
There's historical speculation that a smart TV could connect to an open wireless access point, or more realistically, that it refuses to operate without internet access, perhaps after a certain number of power on hours.
I think this is partially restating "try smarter, not harder" with a lot more words.
I also think it's unkind not to recognize that we have limited time and energy and it's simply not possible to address everything all at once.
IMO the better takeaway is to learn to admit when we're doing that (deprioritizing a problem we don't have the resources to address,) rather than pretending there is no choice, so it occurs to us to revisit the problem if and when there are the resources to do so. My personal approach to this would be to add it a todo list with no assigned due date.
Also, I don't know who the author is talking about, but when I read:
"These are people who could successfully launch a product in a foreign country with little instruction, but who complain that there aren’t any fun people to meet on the dating apps."
I hear someone who maybe isn't valuing romantic relationships but also views admitting that as socially taboo, so they come up with an excuse for why they're not in a relationship. I don't necessarily perceive someone who isn't applying agency to all areas of their life.
With clothes, make it cheap so if you make the wrong decision it's not a big deal.
Also recognize you're engaging in the sunk cost fallacy by keeping clothes you don't actually want, and you're making the world better by allowing it to go to an owner who would better appreciate it.
Some more concrete ideas:
1. thrift stores
2. clothes rentals
3. clothing swaps
4. Buy the cheap version of what you think you might like (if it exists) first before buying the expensive version
5. Don't make your entire wardrobe trendy clothes. Make most of it relatively classic / basic and limit "trendy" to a subset of items.
Inappropriate / unkind, perhaps. I did make at least one solid friend over it.
She saw the post on here, emailed me, and I ended up talking for over an hour with an understanding person who has lots in common with me and gave me some things to think about.
I could have done without the email from someone else who said that "people in Gaza are suffering worse than you and they aren't doing what you're doing, so get mental help."
The absurdist humor of my impending demise making the front page of hackernews and being debated by the denizens thereof is not lost on me.
I was quite moved by the blog post and the brief explanation. That someone found the time to talk to you is good. Even though we don't know each other, I think of you.
Well, sharing someone's deeply personal post on here could be seen as unkind
or inappropriate. I get the sentiment; I can agree with it to a point.
If someone else wrote a similar thing on their blog, I wouldn't submit it here without asking.
But I also think frank, honest, and open discussion of this sort of thing is a net good, which is why I published my post,
and why I don't mind it being here.
If it doesn't already, I suspect training needs to include those no-solution scenarios and reward not overstepping bounds, or else we're going to see a lot more harmful side effects.
reply