drakonka

Thank you for sharing - I'm glad you lived!

On the programming front, I've been working on a little self-hypnosis app for myself that I'll probably go ahead and release. And outside of that I've been finishing up a novella draft.

Mine is: https://liza.io

I'm fine with email but there's also a comment system on the posts that I probably prefer :)

I think this one might be super interesting for its entertainment value, but there has been some strong criticism as to its credibility. For example, here's a pretty comprehensive debunking from The Skeptical Inquirer: https://skepticalinquirer.org/exclusive/the-telepathy-tapes-...

And a statement from Association for Science in Autism Treatment: https://asatonline.org/media_watch/asat-responds-to-the-skep...

Do you think if you're in that kind of situation, where you are dying on a beautiful mountain, you would have it in you to appreciate the beauty around you? I was imagining it to be more like a cold, painful death rather than a nice end where you appreciate the mountain air.

Of course the distinction itself is important in its own way, but I think/hope everyone here fully realizes and accepts that he technically donated as a private individual. You can clearly see it from the headline - it does not say "Mullvad VPN AB is the main financer of the Swedish Örebro party", after all. For those who have an issue supporting Mullvad after this, that is unlikely to be the point of contention. The issue seems to be simply in this case, some people clearly find the private individual's actions objectionable enough to not want to support a product that individual is deeply associated with, profits from (and then funnels those profits into said actions), etc. Fixating on this being done as a private individual just comes off like a bit of a deflection attempt from a pretty straightforward case of individual actions having consequences.

I started doing this after moving to a university town at the end of 2024. Just taking random evening courses that sounded interesting. Having never studied past high school, it's been excellent! My first ever course was on nuclear weapons and disarmament and it was fascinating - we had professors with such a huge amount of experience researching and working directly with disarmament negotiations, policies, inspection missions, etc. Then it was a biology course with the most passionate teacher I've ever met. Last term I took a cosmology course and learned a ton about cosmology from different perspectives. My professors have all been great in the few courses I've done so far, and I don't pay attention to other students so much aside from when necessary for group work.

Mm if I had to actually see myself dying, probably not. If I could just get a letter with the information, as with a short story I read once, probably yes. Assuming the expiration can't actually be changed. If it's not set in stone, that would probably be too stressful to live with.

I sometimes get emails and blog comments like this and always love them. One of my favorites was a comment last year, left on a 15 year old post about building my first gaming PC. I love how the comment said it was a "really fantastic build for the time". Something about "for the time" made me feel so retro :D

I was not actively applying but it ended up being a bit under a month from layoff to offers.

What are your philosophical thoughts on paying a subscription to get your work done?

I've paid for subscriptions to get work done for years. Educational materials, IDEs, source control tools, debugging tools, etc. I don't think the fact that you're paying for a subscription is the real potential worry here, right? It's more about the effects of using this particular tool.

Both of those questions have everything to do with feelings, no?

I have observed this in myself when I began to over-leverage AI in my workflows. I've since become more deliberate with what kinds of tasks I will use it for, although I still slip up.

With writing:

Things like brainstorming a plot line for a book with a custom GPT or Claude project that has all of my prior books in its knowledge? Works great.

Things like asking it to write a paragraph or chapter for me - I can rapidly feel my own writing skill, motivation, vocabulary, and ability to grasp/remember the resulting plotlines deteriorating. I don't use it for that anymore.

With studying:

I've been taking a couple of evening uni courses and the thing I found so great is that I've been forcing myself to think through the problems, and take my own notes in every lecture. I may then still get ChatGPT to help explain and reason through some of the concepts with me. And I have it review and 'grade' my assignments. But I refuse to ask it to start drafting answers.

With programming:

This one is tougher. When I am not very personally invested in a problem or codebase it becomes too easy to offload more parts to Claude, and when the company encourages 'vibing' to speed up velocity and you're reviewing and writing a higher influx of lower quality PRs, investment goes down. I still sometimes catch myself committing solutions I only _mostly_ grasp and the rest is hand-waving. A big part of it is a work culture thing.

For my own projects I make sure to understand and have a back-and-forth with the planning agent for each task, or write the first plan myself to go off of. When it comes to producing the code, I have to admit it is much easier to properly review parts of the codebase I am extra interested and knowledgeable in (backend in my case). The frontend I'm less well versed in and also admittedly less interested in, so I do sometimes fall into the trap of "Ehh it works, just commit it" with the goal of doing a thorough quality pass before actual release.

With all of the above, I can feel my ability to think, plan, reason, focus (and my vocabulary) suffer if I go over the line too much into agent offloading. For me keeping that balance is as much about maintaining my own long-term brain health as it is about producing good output. I imagine younger people growing up with AI today won't even know what that more capable (in my opinion) brain state feels like - to them, the AI-using brain will be the norm.

I'm not sure what the take-home of this message is since you can replace "AI" with just about anything... "If you use a keyboard to do your work, you can be replaced by someone else using a keyboard to do your work." Sure? You can always be replaced by someone/something that can do your job better.

Great idea, done.

I fixed the second one normally because that's just embarrassing. I guess at least we know it was really written by a human? ;)

Hahaha I missed that completely, sorry.

...Now I kind of don't want to fix it :D

For me the small screen is fine as long as I can maintain the mindset of this being a _drafting_ tool, not an editing or structuring tool. I generally don't need to see more than a couple of sentences at a time, and my drafting process is almost stream-of-consciousness as I just focus on getting the words out. If I needed to navigate more to edit or regularly reference prior sections, the small screen would be a hindrance.

I chose the keyboard I wanted :)

I use the Onyx Boox Palma for a portable eink drafting setup. It's worked pretty well. I wrote about it here: https://liza.io/portable-writing-setup-with-onyx-boox-palma/

I post a lot of notes on my personal blog. From lectures and such. I use AI to proofread the notes I took during the lecture. Other than that it outputs them in exactly the same lecture-dump format that I take them in, and then I edit them into a structure that makes more sense and add further reflections etc myself by hand.

For me, the point of taking and posting the notes is to solidify the lecture in my mind, think about it, etc. Using AI for too much of this process would defeat the purpose.

My team was recently laid off, myself included. Before formal termination my employer had to negotiate with my union and run everything past them - from a potential exit deal (which I opted not to take) to the whole layoff procedure and new organizational structure. This is to make sure it was all by the books and I was getting everything I was entitled to. We have no collective agreement, but any employee who is individually part of a union has this right.

Some employers (completely non-maliciously, I should add!) simply don't know all the relevant labor law either, like summertime vacation entitlements, how the employee is allowed to search for other work during the notice period, unemployment insurance stuff, etc. It made a stressful situation much better for me knowing I have someone who knows this checking everything and advocating for me.

I keep seeing people say "unions won't work" in other countries. Who knows, maybe they're right. I'm just glad I don't live in one of those, I guess.

That seems like a really tenuous connection for him to make if that's what he's doing. I find it difficult to believe that ambitious people outside of a certain niche would refuse to come to the US because of perceived lack of AI progress. They'd refuse/are refusing to come because of the US's increasing hostility to outsiders, cutting of research funding, and subpar living conditions.

Not about sleep learning but lucid dreaming in general: I have long been puzzled and disappointed that we are not pouring more research, interest, and funding into controlling the induction of lucid dreams. There are ways to learn to do it now, but they are slow and unreliable. Gadgets exist but they're fringe - there is no great lucid dreaming movement like there is with longevity, for example.

I would have thought in a society where we want a gadget or magic pill for anything and everything, our interest in this would be through the roof. You can live a whole other lifetime in a dream, either your own or of some other character or world you step into. You can replay the past see the future, countless times. Controlled lucid dreaming seems like the closest we will likely ever get to immortality. Why aren't we more bullish about facilitating it?

I wonder if you would ask the same thing about any number of apps - like fitness trackers, mood trackers, supplement trackers, online diary apps, task trackers, etc? You don't even need a notes app - you could just carry a notebook around or email notes to yourself.

As for why people may want to track menstrual cycles specifically, it is because bodies can be greatly influenced by what phase of the menstrual cycle we are in. From regular physical and mood changes to disorders like PMDD. The different parts of the cycle can also impact ideal exercise and even food choices for some. There are women and couples who gain insights (and often useful predictions) into how their moods coincide with menstrual phases, and that is much easier to track in a dedicated app designed to do so (which can also flag cycle irregularities, bleeding variation, or other changes), just as with other purpose-built applications. All of that is before we even get to the whole fertility tracking thing. One such app is a certified birth control method in my country. Tracking periods in a notes app is not.

Half asleep definitely doesn't sound like good quality, restorative sleep.

I have been thinking about this kind of thing recently. I've got a hobby project that generates some AI content for the user and trying to figure out the most fair way to deal with the output being just plain bad. I haven't run into this case myself yet in testing after many quality passes to make the generation robust, but have no doubt that at scale there would be junk output at some point. Users would likely be non-technical and pay per-generation, not on a subscription model. So I would like to find a way to:

- Define some threshold for bad output

- Detect when a piece of output meets that threshold (vs just maybe not being what the user expected, which in my case is just fine)

- Refund the user credits so they can generate again

Text output is relatively easy to evaluate to some base threshold of quality in the generation process, but the final output is not text... it becomes harder.

Each failed generation would be very disruptive to the user on its own (in the scope of the app's purpose), so I'm also considering offering them an extra discount on their next purchase (in addition to the credit refund).

Do I get users to report generations they consider bad and then review them somehow? Do I try to auto-detect bad output before the output is delivered to the user? Probably a mix of all of the above... while attempting to mitigate the potential for abuse (people making dummy generations and then reporting them 'just to try it out', or try to game the system to get multiple free generations). Maybe I'd have to have some sort of time window for reporting a junk generation, and a max "use" count that flags if the user actually took benefit from the output before reporting it...

I guess this turned into a bit of a brain dump.

I think it is already pretty widely recognized that caffeine can disrupt sleep taken even as early as 6 or 7 hours before bed. I usually don't drink coffee or caffeinated tea after 12 for this reason. Caffeine also has many other known benefits, possible beneficial effects on all-cause mortality, etc, and I'm not sure if we have any research showing the same benefits coming from paraxanthine. Seems like potentially a bit of a waste just to be able to get the stimulant effect a couple of hours closer to sleep time.

I think I got a really cool wallet and notebook from Hand & Sew through Massdrop many years ago. The wallet is long gone but the notebook is still around with a wonderful patina. It used to be a really cool site for unique items. I stopped using it when they had a falling out with Input Club and it seems like they've just gone downhill since then.

Reminds me of a talk I went to in 2018 about rebel agents, in which the speakers talked about some ongoing work in this area and gave some good examples of physical systems that we might _want_ agent rebellion (e.g., a delivery drone is instructed to take a certain route, but the operator instructing it may not be fully aware of the situation or the specific obstacles in the drone's way (or maybe even all of the drone's underlying goals). The drone may then choose to 'rebel' and deviate against the operator's instructed flight path).

They also talked about the importance of explanation (on the agent's part) using theory of mind regarding why it rebelled. I took some notes at the time and put them here: https://liza.io/ijcai-session-notes-rebel-agents/

It makes me so happy every time I think about that - how my cats have never had to know anything other than utter love and safety.

In actuality he's been through some really rough times at vet hospitals as a kitten (and that's why he reacts differently with vets), so it is an extra relief that he came out of that so well-adjusted at home.