mathewsanders

This makes me think of a tool from semiotics called the Greimas square where you can have opposing concepts e.g. A and B (ugly & beautiful, for & against, legal & illegal).

At the surface level they can appear as binaries, but the negation of A is not equivalent to B and vice versa (e.g. illegal is not equivalent to not-legal) and encourages the consideration of more complex meta-concepts which at surface level seem like contradictions but are not (both beautiful and ugly, neither for or against).

--

Others have pointed out that English speakers do have the capacity, and do use these sort of double negatives that allow for this ambiguity and nuance, but if you are an English-only speaker, I do believe that there are concepts that are thick with meaning and the meaning cannot accurately be communicated through a translation - they come with a lot of contextual baggage where the meaning can not be communicated in words alone.

--

As a New Zealander who's lived in the U.S. for the last 15 years, I've realized in conversations with some native Americans where despite sincere (I think) efforts on both sides, I've not been able to communicate what I mean. I don't think it's anything to do with intelligence, but like author hints how language shapes how we think and therefore our realities.

--

I've never found poetry to be interesting, but recently I've come to appreciate how I think poets attempt to bypass this flaw of language, and how good poets sometimes seem to succeed!

Something I think would be interesting for model APIs and consumer apps to exposed would be the probability of each individual token generated.

I’m presuming that one class of junk/low quality output is when the model doesn’t have high probability next tokens and works with whatever poor options it has.

Maybe low probability tokens that cross some threshold could have a visual treatment to give feedback the same way word processors give feedback in a spelling or grammatical error.

But maybe I’m making a mistake thinking that token probability is related to the accuracy of output?

This press release says it will be available “starting today” through developer program https://www.apple.com/newsroom/2025/06/apple-supercharges-it...

I saw an ENT for the first time earlier this year and was shocked that each visit (less than 5 mins) got billed at around $1500 per visit.

I had a problem with my carrier as well with porting my number over. Attempting to interact with their customer service was a painful loop where nothing was resolved.

After a week I submitted an FCC complaint online (it was very straightforward) and issue was resolved in 24 hours.

During the start of covid I was considering buying a pulse oximeter and it annoyed me that some listings on Amazon were using “FDA approved” in listing and logo and I found it was easy to report them to FDA and their listing was taken down.

One time I was frustrated that a large and popular NYC-based physical store was charging sales tax for clothing under $110 (in NYC clothing and shoes under $110 have 0% sales tax) and I tired reporting to a state authority but I never even got an acknowledgment that complaint was received :/

I have to give a shoutout to New Zealand Marmite which is totally different to British Marmite (I think that NZ marmite has some sweetness as well).

Somehow i grew up in a mixed household that had both because my mum preferred Vegemite and I prefer Marmite.

In NYC it’s fairly easy to get Australian Vegemite but sadly impossible to find NZ Marmite, instead I do a bulk order every couple of years so that I have a stockpile :)

I live in a smallish apartment building with 50 apartments and in the basement we have a little building community library where people put books/DVDs/Blu-ray disks that they’re done with. Last weekend I grabbed The Goodfellas and The Dark Night and will put them back when I’m done.

I think it would be really cool if there were ways for people to share their physical media because I don’t have the room to maintain a big media library, and also don’t have the energy to rip and store locally.

I also want to add that I’ve changed my streaming behavior- I will subscribe and immediately cancel the subscription so that it expires after a month so that I don’t end up with a bunch of active subscriptions that I’m not actively using.

When I do subscribe I always pay for the more expensive ad-free versions but recently I couldn’t get anything to play on Paramounts streaming service. After some trial and error I found that their “ad-free” service won’t run with my blocker running on my router and I needed to allowlist some ad services for it work. That’s pretty annoying.

Maybe you’ve read this already (I still have it in my stack to read so can’t give any personal review) but ‘Determined’ by Robert Sapolsky sets the argument that we don’t have free will and that the justice system is flawed because of that.

https://www.newscientist.com/article/2398369-why-free-will-d...

Nathan for You S3.E6 also explored this idea as a hotel amenity for sexually active parents.

It’s a really interesting product that I’m not going to buy!

Declarative interfaces that allow you to describe what you want and use agents to go out to different services and chain them together is a cool idea:

I don’t want to spend time using dozens of different apps with different (often poorly designed) interfaces.

Having a push to talk hardware button instead seems less clunky than a “hey siri” key phrase (I use Siri dozens of times a day but unfortunately ‘raise to talk’ feature on Apple Watch has never worked well for me).

I’m curious how their LAM works with interfaces being updated- if they need to retrain with UI updates or if it’s flexible enough to be stable with UI changes and new features etc.

I currently use ChatGTP sessions to dive into various topics I’m interested in, and explore ideas- I do like the idea of dedicated hardware that would allow this, but it’s something I imagine I’d keep on the coffee table at home, I don’t want to get a dedicated SIM and data connection or carry around another device.

They’ve raised $30M and I wish them well, I hope they survive and have me as a customer in the future.

> NASA management was concerned about losing the 400,000 workers involved in Apollo after landing on the Moon in 1969

400,000!

It’s probably more due to coastal subsidence than rising sea levels. My favorite is when glaciers retreat there is a rebound effect from pressure which can cause other local land to sink as a sort of equilibrium.

I think in the Mediterranean coastal subsidence is mostly due to plate tectonics. I remember reading something about how myth of Atlantis might be due to a real life city that sunk as a result of an earthquake.

I’ve been biking around nyc boroughs for over a decade now and I still follow directions from mapping apps!

I’ve found Apple Watch to be okay (they give instructions on directions via haptic feedback) and AirPods with spoken directions to be even better.

I wear glasses too, but for directions I think I’d prefer haptic and/or audio even if visual was an option just for safety purposes.

The one pain point I have found with biking directions is that sometimes the directions come way too late to be useful (maybe I bike too fast haha).

I flip-flop between optimism and pessimism on this and I’m in a pessimistic cycle at the moment.

I can also imagine the machinery of capitalism maybe working (I like the idea of declaring a ‘war on emissions’ popping a war revenue act for the highest earners and creating a new industry focused on reducing emissions).

But two things are fueling my pessimism: the first is that as a culture we’ve grown so used to a lifestyle (and economy) based on unsustainable use of our resources. I just can’t see a world where we take a step back from any of the modern conveniences we’ve gotten used to hanging.

I guess related to that: The free market is the most effective model, but that comes at cost of no protections for workers and resources, and that rewards exploitation.

I used to hope that I was wrong, now I mostly just hope that the worst results don’t happen in my own lifetime.

That's fantastic! I was recently doing some research and learnt that with ART people who are HIV+ can actually expect to live a little longer than people who are HIV-. I think it's because they have more regular check ups to make sure that treatment is being maintained.

I guess I don't fully understand U.S. healthcare system (I'm an immigrant from a country with universal healthcare, and this is the first time in my life that I've had a medication prescribed other than short-term antibiotics).

I pay a $10 co-pay for generic with my insurance, but when I pick up from pharmacy the receipt says "you insurance saved you $1888" (you're right it's not ~$2500/month - I just checked and that's a different prescription I have that I get filled at the same time).

> They can take pre-exposure prophylaxis (prep), a kind of drug that reduces their chance of contracting it by 99% or so. This comes as a daily pill, and is popular among gay men in rich countries.

The article takes a global focus (which makes sense since most new HIV diagnoses happen outside of the United States), but even the United States has over 30,000 new HIV infections per year.

In 2021 feds mandated that insurance companies must include PrEP as free preventative care, but that doesn't help the most at risk who are young and may not be insured (30-day supply of generic PrEP costs ~$2,500, and every 3 months people must have doctor visit and bloodwork to check HIV and STIs, and for potential liver and kidney damage).

In 2022 a federal judge ruled that they can not require Christian companies to cover PrEP as this violates their religious rights under federal law[1]. This will likely make it's way up to the Supreme Court, and there's a high risk that as a result people in the United States will then need to pay out of pocket for PrEP.

[1] https://www.axios.com/2022/09/07/court-hiv-prep-requirement-...

Those smaller controlled burns help prevent organic matter from building up that can lead to more devastating uncontrolled fires. The national parks service has been incorporating those methods into their fire management practices.

And driving bison of cliffs were a sustainable hunting practice - it was habitat loss due to farming and commercial hunting that led to their population decline.

Gematria is featured a lot in scott alexander’s Unsong (serialized fiction that was published in via blog) http://unsongbook.com/

Very quick summary is that Kabbalah is real, and corporations set up content farms to brute force search for names of god so they can copyright and sell them.

It’s a really fun read (author probably also has a hacker news account, so if you see this comment thank you for writing such a fun book).

+100

> Because we don't don't have any memory of what it was like to be in these states.

I want to add an anecdote from about 5 years ago when I fainted: there is a tiny moment as I was regaining consciousness that I had an experience that is unlike anything I’d ever experienced before (and words really fail to describe it) there was no sense of time, or even presence- it was like pure observation. It wasn’t the blackness nothingness that I think is associated with being unconscious, instead it was the opposite- it felt like hundreds of images, sounds and thoughts all layered on top of each other- it was very ‘noisy’. That might sound stressful, but I don’t remember feeling anything at all.

It was only when some of that noise faded away that I experienced a sense of ‘self’ again and a moment later I formed the idea where am I? And then I opened my eyes and everything was back to typical conscious experience (although feeling a little disoriented).

Biological neural nets have a benefit of being able to run massively parallel, I know basically nothing about silicon hardware, but I’m assuming that attempts to model hugely parallel systems like a brain on a serial architecture is going to be magnitudes more inefficient.

I do remember years ago reading an a response to a question someone asked about why can’t we just build chips that run in parallel and there’s the constraint of printing on a 2D surface that limits how far you can take that idea.

> the oppressed themselves don't understand it well enough to avoid oppressing others

I can’t comment too much because i’ve had a copy sitting unread in my pile of books to read for over 6 months now, but if anyone is interested, I think this is the exact topic of Paulo Freire’s ‘ Pedagogy of the Oppressed’ https://en.m.wikipedia.org/wiki/Pedagogy_of_the_Oppressed

The author claims to be rational, but they don’t volunteer any alternative hypotheses to demonic possession that they’ve considered, eg speaking ‘unknown’ languages could be exposure that they had as a child and had since forgot (still pretty remarkable, but less remarkable than paranormal entity), ‘knowing’ facts about strangers could be just a guess and there’s a sort of survivorship bias where only the correct guesses are documented.

Levitation is harder to explain, but also the author does not claim to have witnessed that themselves so the easier explanation is that it’s not an accurate report of whatever occurred.

Having said that, I think it’s important to stay curious about these claims. A close friend and myself experienced a physical phenomenon, and many years later I’m still don’t have any plausible explanation for what it might have been.

Our company is mandating a 3-day in office starting soon.

I’m a bit of an exception because I’m in a city that only has a couple hundred employees, but when I do return to the office, it won’t be to collaborate with teammates: I will be joining zoom calls with people in other cities (which was my day-to-day pre-pandemic). Pre pandemic I would travel every 2-weeks to company HQ to be in-person for demo with stakeholders and sprint planning, but even then many people would join via zoom anyway.

I don’t think return for office has anything to do with productivity and more about anxiety around investments in commercial real estate.

For Manhattan they’re giving cost of living for 1 adult as $53,342 which is way way off.

They are allocating $2000/month for housing (avg 1 bed in Manhattan is around $4000) and $12.50 for food per day (I guess possible if you’re eating rice-and-bean type meals).

This living wage is 4x povity wage, so I guess they’re expecting povity line to be finding places that rent for $500 and being able to budget for $4 for food each day.

Thankfully we voted in last years local election to have a true cost of living index for nyc calculated each year. http://truecostofliving.org/

For 2023 they estimate north Manhattan as $123,688 And south Manhattan as $151,723

Shocking how far off the calculations are given that I assume a resource like this is used for policy decisions.

I had an order from Floyd (online furniture retailer) and on day of delivery the delivery company said the order was damaged and to contact Floyd for replacement.

Amazing run around where every week I would email Floyd customer support, they’d said they would order a replacement in a ‘few days’.

Six weeks of that and I submitted a complaint though BBB and I immediately got a reply that the items were now out of stock, and would be likely in stock in 6 months or so.

I ended up accepting the damaged items (at full price, no compensation) rather than wait so long.

A few years ago online-only startups distinguished themselves by providing better than brick-and-mortar customer service but I see this has flipped. I think this is what Cory Doctorow calls enshittification.

The idea that people’s minds can be so different is something I didn’t realize until a couple of years ago and has been a lightening bolt moment for me in better understanding myself and others!

I believe I have aphantasia, and the idea of being able to visualize things on demand seem so strange to me that it’s hard for me to even imagine (isn’t it distracting?).

When I heard things like “count sheep to fall asleep” or “imagine yourself on a beach” to relax, I had assumed this was a figure of speech, but apparently people are doing this literally!

I’ve spent some time trying to practice visualizing things mentally and sometimes if I relax and try not to think about it I’ll get a glimpse of something (last time I tried I got something like looking at a forest from the sky) but the moment I try and focus it disappears in a snap, so I feel like it’s closer to dreaming than visualizing something mentally.

Related to this, I have plenty of audible conversations in my mind (either with myself or other people) and I’ll often have a song playing in the background of my mind (although music seems to be limited to just remembering things I’ve already heard) apparently some people can’t imagine sounds or voices which also seems equally bizarre to me.

I was so excited to try this because I have a book that I’ve not been able to locate (have tried my own searches, Reddit and online forums) but ChatGPT gave me a lot of hallucinations and listed out real books and incorrectly claimed that all those books had the plot bit that I’d remembered.

Policy is what is said, and that’s often a huge gap between what actually happens.

Tumblr had a huge content moderation issue where (based on my personal experience) I’m guessing there are incentives in place around flagging explicit content. I’ve never posted or reblogged anything that would break this policy, but every single post I make on my main blog is automatically flagged as explicit/adult content (eg a photo of my cat) and going though a content appeal I’d say 50% the human-reviewed content appeal fails and my post gets hidden.

I ended up creating a separate blog that I now post content to, and then reblog to my main account.

It’s impossible to get a hold of any support to try and resolve.

Despite non-existent support and increasingly buggy apps, it’s still the best online community I’ve found and I love it :)

Tumblr is the only platform that I’m still active on- I’ve had an account since 2009-ish. It’s a really fun community to be active with and the only one that hasn’t attracted a majority of insufferable people.

It’s great to see the Richard Rumelt format of strategy with diagnosis/guiding principles, I’m a fan of this format but it also feels a bit like strategy-by-committee and just a laundry list of things.

Reading though this I get the sense that no one at tumblr has great insights around the pain points of their own product. I think they need to take a step deeper in their diagnosis and ask “how did we end up with a platform that’s so hard to use?” Otherwise they won’t make progress on the actual underlying issues.