rebeccaskinner

Jellyfin will let you set up alternate metadata providers and select between different episode ordering if available (https://forum.jellyfin.org/t-resolved-dvd-order-instead-of-a...). For tv shows especially TVDB will tend to have alternate orderings most often, and TMDB will occasionally have alternate orderings available.

I don’t think pricing is the only problem with theaters. Especially over the last 15 years or so they’ve been increasingly competing with alternative ways of watching movies, and for a lot of people watching at home wins at any price.

Fancier theaters like the Alamo draft house seem to be trying to complete with watching at home in some ways, but for the most part theaters seem to just be doubling down on the parts if the experience that were already decisive- namely getting louder and adding bigger screens. That might tempt the people who already like what theaters have to offer into going slightly more often, but I think it’s made even more people stop going at all.

It is pretty riddled with ads though. By default the Home Screen shows the tv all and it’s completely crammed with ads.

I sometimes talk with ChatGPT in a conversational style when thinking critically about media. In general I find the conversational style a useful format for my own exploration of media, and it can be particularly useful for quickly referencing work by particular directors for example.

Normally it does fairly well but the guardrails sometimes kick even with fairly popular mainstream media- for example I’ve recently been watching Shameless and a few of the plot lines caused the model to generate output that hit the content moderation layer, even when the discussion was focused on critical analysis.

I shared the story as I remember it. Memory is imperfect. It's been years since I deleted my account, and I don't have the luxury of access to server or moderation logs.

What I do remember unambiguously is being an active member of the site, contributing regularly and in good faith, being accused of spamming, and the general feeling of hostility that I got from the site.

I was a pretty active member in the comments for a long time and left a few years ago after getting chastised by a moderator and accused of spamming for sharing a link to a blog post I had written, even though the content was purely technical, not promoting any product, and does not contain ads or monetize content in any way.

My impression is that the site was actively looking for any possible reason to remove people from the platform. It’s their site to moderate as they wish, but that’s not a community I want to continue participating in.

The only acceptable number of ads is zero.

I’m finishing up Haskell Brain Teasers (https://pragprog.com/titles/haskellbt/haskell-brain-teasers/)

It’s much shorter than my first book, Effective Haskell, and leans more advanced, especially toward the end. Although the format is puzzle focused I’m trying to avoid simple gotcha questions and instead use each puzzle as a launchpad for discussing how to reason about programs, design tradeoffs, and nuances around maintainability.

I worked on a foveated video streaming system for 3D video back in 2008, and we used eye tracking and extrapolated a pretty simple motion vector for eyes and ignored saccades entirely. It worked well, you really don't notice the lower detail in the periphery and with a slightly over-sized high resolution focal area you can detect a change in gaze direction before the user's focus exits the high resolution area.

Anyway that was ages ago and we did it with like three people, some duct tape and a GPU, so I expect that it should work really well on modern equipment if they've put the effort into it.

Although thus isn’t directly related to the idea in the article, I’m reminded that one of the most effective hacks I’ve found for working with ChatGPT has been to attach screen shots of files rather than the files themselves. I’ve noticed the model will almost always pay attention to an image and pull relevant data out of it, but it requires a lot of detailed prompting to get it to reliably pay attention to text and pdf attachments instead of just hallucinating their contents.

I’ve had similar experiences where AI saved me a ton of time when I knew what I wanted and understood the language or library well enough to review but poorly enough that I’d gave been slow writing it because I’d have spent a lot of time looking things up.

I’ve also had experiences where I started out well but the AI got confused, hallucinated, or otherwise got stuck. At least for me those cases have turned pathological because it always _feels_ like just one or two more tweaks to the prompt, a little cleanup, and you’ll be done, but you can end up far down that path before you realize that you need to step back and either write the thing yourself or, at the very least, be methodical enough with the AI that you can get it to help you debug the issue.

The latter case happens maybe 20% of the time for me, but the cost is high enough that it erases most of the time savings I’ve seen in the happy path scenario.

It’s theoretically easy to avoid by just being more thoughtful and active as a reviewer, but that reduces the efficiency gain in the happy path. More importantly, I think it’s hard to do for the same reason partially self driving cars are dangerous: humans are bad at paying attention well in “mostly safe and boring, occasionally disastrous” type settings.

My guess is that in the end we’ll see less of the problematic cases. In part because AI improves, and in part because we’ll develop better intuition for when we’ve stepped onto the unproductive path. I think a lot of it too will also be that we adopt ways of working that minimize the pathological “lost all day to weird LLM issues” problems by trying to keep humans in the loop more deeply engaged. That will necessarily also reduce the maximum size of the wins we get, but we’ll come away with a net positive gain in productivity.

Looking at my own use of AI, and at how I see other engineers use it, it often feels like two steps forward and two steps back, and overall not a lot of real progress yet.

I see people using agents to develop features, but the amount of time they spend to actually make the agent do the work usually outweighs the time they’d have spent just building the feature themselves. I see people vibe coding their way to working features, but when the LLM gets stuck it takes long enough for even a good developer to realize it and re-engage their critical thinking that it can wipe out the time savings. Having an LLM do code and documentation review seems to usually be a net positive to quality, but that’s hard to sell as a benefit and most people seem to feel like just using the LLM to review things means they aren’t using it enough.

Even for engineers there are a lot of non-engineering benefits in companies that use LLMs heavily for things like searching email, ticketing systems, documentation sources, corporate policies, etc. A lot of that could have been done with traditional search methods if different systems had provided better standardized methods of indexing and searching data, but they never did and now LLMs are the best way to plug an interoperability gap that had been a huge problem for a long time.

My guess is that, like a lot of other technology driven transformations in how work gets done, AI is going to be a big win in the long term, but the win is going to come on gradually, take ongoing investment, and ultimately be the cumulative result of a lot of small improvements in efficiency across a huge number of processes rather than a single big win.

It would be much less interesting than the actual chat histories. My experience with chatGPTs memory feature is that about half the time its storing useful but uninteresting data, like my level of expertise in different languages or fields, and the other half it’s pointless trivia that I’ll have to clear out later (I use it for creating D&D campaigns and it wastes a lot of memory on random one-off NPCs).

Maybe it’s my use of it, but I’ve never had it store any memories that were personally identifiable or private.

I remember back in the gnome2 days there was still a lot of fragmentation. Gnome, KDE, WindowMaker, AfterStep, Enlightenment, ratpoison.

Linux has always appealed to tinkerers and that was always going to lead to some amount of fragmentation. I don’t think it’s a bad thing necessarily. For all of the complaints about it, systemd has unified a lot of things that used to be handled through desktop environments and made things less fragmented as a whole.

The biggest issue to me is that beyond and impossible aren’t just making replacements that are worse than meat, they are making things that are worse than the alternatives we already had.

A beyond burger might be more like meat than a patty made from beans or lentils, but it tastes worse and has a worse nutritional profile. Beyond chicken isn’t even all that similar to chicken and it’s a worse substitute than seitan for something like wings.

join has the type `m (m a) -> m a`. That's the thing that really shows off the monoidal structure. People normally implement monads in terms of bind, but you can easily define join in terms of bind for any Monad: `join ma = ma >>= id`. So really, as long as you have a lawful instance of Monad written with bind the existence of join is your proof.

Unfortunately even without engine noise people will still install gigantic subwoofers in their car and roll around rattling windows. This has consistently been the biggest problem with my sleep for years, and nothing I've been able to do (including trying to move to different locations) has helped because they are so incredibly loud and prevalent these days.

Several years ago I (briefly) worked at a startup that was trying to do this for publishing (but has since pivoted into generic ad-tech). My impression at the time was that most publishers weren’t onboard. True or not, they seemed to think if you’d pay a penny for an article then you might but a subscription and so they want you to make an account, want your contact info so they can send you spam, etc.

The other issue is that big name publishers saw micropayments as eating into their subscription revenue and weren’t interested, but without them it was hard to put together a compelling enough bundle of sites to overcome the signup friction for users.

I still think it’s a good idea but I don’t see how you overcome those obstacles.

I’ve been lucky enough to work on a number of teams in my career that are full of brilliant high performing people who are also kind, empathetic, and focused working together to deliver something to customers. These days I’m lucky enough to lead a team like that.

Counter to the article, in my experience these are the hardest teams to manage well because organizations typically aren’t set up to deal with them. Larger companies tend to lean into standardization and making things accessible to average engineers in ways that make high performing teams less effective and often demotivated.

I think this is an overly cynical approach that assumes that you can’t invest and grow people into exceptional engineers. In my experience you can if you are willing to invest in it, and the long term benefit of having more high performing engineers who aren’t being restricted from doing good work outweighs the cost of training and growing people.

I don't really remember when they started, but they now include 1-2 prerolls for other Apple TV+ content. You can scrub past them, but you can't disable them.

yeah, the pre-rolls for the other shows are exactly what I'm talking about. Those are ads. And they are only skippable if you scrub forward past them. You are forced to deal with them.

Apple TV+ / movie-streaming services (without ads)

Apple TV+ doesn’t have an ad-free tier. They have a tier they call ad-free where they still force you into pre-roll ads for apple products. Unfortunately that’s the same for most streaming services that claim to be ad free.

Don’t forget recursion schemes. The last example in the article was just asking for a hylomorphism.

While it's possible to bake in this particular inductive bias (repetitive actions might be useful), they decided not to (it's just not that interesting).

What's interesting to me about this is that the problem seems really aligned with the research they are doing. From what I can tell, they build a system where the agent has a simplified "mental" model of the game world and it uses to predict actions that will lead to better rewards.

I don't think what's missing here is teaching the model that it should just try to do things a lot until they succeed. Instead, what I think is missing is the context that it's playing a game, and what that means.

For example, any human player who sits down to play minecraft is likely to hold down the button to mine something. Younger children might also hold the jump button down and jump around aimlessly, but older children and adults probably wouldn't. Why? I suspect it's because people with experience in video games have set expectations for how game designers communicate the gameplay experience. We understand that clicking on things to interact with them is a common mode of interaction, and we expect that games have upgrade mechanics that will let us work faster or interact with high level items. It's not that we repeat any action arbitrarily to see that it pays off, but rather that we're speaking a language of games and modeling the mind of the game designers and anticipating what they expect from us.

I would think that trying to expand the model of the world to include this notion of the language of games might be a better approach to overcoming the limitation instead of just hard-coding the model to try things over and over again to see if there's a payoff.

I think most people care about some things. Most people don’t have the capacity to feel strongly about every issue, but there are some overarching ones that tend to hold along political lines, and people will tend to have pet issues as well.

I think there are parts of the policy that people wouldn’t be able to agree on because of differences in values.

To look at another example, some people view the purpose of prisons as being primarily for causing suffering and to punish people. Other people don’t care much either way about suffering and see prisons as a way to remove people from society. Some people think the purpose of prisons should be rehabilitation, and see suffering as practically counterproductive. Some people don’t believe that if the state is taking someone’s freedom they have an ethical obligation to minimize that persons suffering. Some people don’t believe in the concept of prison at all.

There are a lot of views there, and while you might be able to get some of the people with differing views to agree on policy some of the time, the goals are significantly different and that’s going to be a significant obstacle in shaping a meaningful policy in all but perhaps a few isolated cases.

pro-choice, pro-LGBT, and pro-immigration may very well vote Republican because they work in the US Automotive industry, and so do their friends and families and people who they care most about.

What you care most about is a statement of values.

I reject this idea, someone voting for the "least worst candidate" does not wholly endorse everything they stand for

The thing about values is that they don't just capture the notion of what we thing is right or wrong, but also which things we value over other things. In an extreme case, two people can agree on 10 out of 10 different ideals or ethical stances and still have different values and support different parties because of how they rank those things.

In that case who you think is the "least worst" is also a reflection of values, as is declaring both sides to be the same, or opting out altogether. They all represent both what things you value and how much you value them.

Don't bother with Those People, They Have Different Values, They Aren't Like Us

I didn't say that you shouldn't bother with people. I said that discussing _policy_ is not useful if you don't agree on _values_. It's the wrong level of abstraction. To put it in a plain analogy: discussing the best route to get to your destination isn't useful if you don't agree on where you are going.

If you want to engage with someone with different values, then the values are where you need to start. If you want to engage with someone on the best way to get somewhere, you need to start by making sure you both agree on where you want to go.

For all of the author's bloviating and self-congratulating navel gazing, the article manages to largely overlook values, the only mention of them being to dismissively reduce them to irrational tribalism.

In truth, values and ethics are fundamental to effectively discussing politics. After all, all political decisions are ultimately about how we want to shape the world that we as humans live in. There can be no agreement about economic policy without a shared understanding of the ultimate goal of an economy. No agreement about foreign relations without a shared understanding of the role of nations as representatives for groups of humans, and how we believe one group of humans should interact with another group of humans through the lens of nations.

For the last 20 years at least, the leadership of the two main political parties in the US have largely invested in messaging around the values that they represent. The policies are different too, but over time we've gone from a world where there were at least some cases where the two parties had different policies for how to reach the same goals, and into a world where the parties policies are aiming to realize fundamentally different visions of the world, based on fundamentally different values.

In this world, asking "who did you vote for" isn't a matter of tribalism, but it is a (good) proxy for asking someone "what are your values". If you discover that someone has completely different values from you, then discussing policy isn't going to be useful anyway, because there's no way you'll agree on a single policy when you have different fundamental values.