psandersen

Great point, the incentives and mutual interests are definitely there. I think it comes down to how ideological the opposition to climate change is in Aus... I.e. are they completely committed to not doing anything effective to mitigate it that might break their narrative around climate change. I'm pretty jaded after the NBN...

Thanks for the link, already ordered the other sensors. Looks like SGP30 isn't quite as easy to use with ESPHome right now, so something to upgrade to within a year. Its really amazing how great the ecosystem is with ESPHome, I was always a bit scared buying the chips and getting really DIY with the likes of Arduino but this is no more difficult than most other tech.

Whats the chance that $500M is actually to help the reef and not just used for corruption or forgotten about until the next election? I have zero faith in the current Australian gov doing anything to mitigate climate change...

Otherwise this is a great idea!

I personally believe morality and self interest strongly overlap over the long term, but short term they are largely independent conditional on the probability of getting caught...

Humans might have too short lifespans, memories and limited rationality for the long term benefit of morality to be strongly in our individual self interest though... one of the possible benefits of anti-aging and cognitive enhancement tech is it might incentivize us to be more moral all other things equal as a side effect.

On a related note, I've been looking for a good sensor that is hopefully plug & play and reliable for integration with home assistant. Is something like https://www.banggood.com/Geekcreit-Nova-PM-Sensor-SDS011-Hig... really the best way to go about it rather than an Awair? Would be keen on VOCs and CO2 as well.

edit: Looks like esphome with the following low-cost sensors is the way to go: * MH-Z19 for co2 (or CCS811 CO_2 for VOCs as well) * SDS011 for particle counts

What human bias would a knowledge seeking AIXI ( https://en.wikipedia.org/wiki/AIXI ) agent have? AIXI is incomputable but that doesn't mean unbiased approximations would have any resemblance to any human-provided/designed priors after some years/centuries.

Pricing in externalities are the only good economic argument for tariffs I'm aware of...

Carbon tax + tariffs in proportion to the lack of carbon taxes would be the most logical way of setting up a market-based solution to climate change.

The problem is people who would pay to avoid ads are those worth targeting... I'd love if a single service could let me opt out of all tracking and ads internet wide for a fixed fee, say $50-$100 a month.

I'm getting tempted to go the LineageOS with MicroG route on my Pocophone (need to confirm it works first).

Does this mean I can use the Play store and apps such as Spotify without much issue?

You actually have to make it deliberately slow if Google's js is disabled? that seems a little malicious...

It is a huge amount of money. Though the real question is the feedback loops UBI creates IMHO... The money will be injected straight back in the economy as people are still living paycheck to paycheck at that wage, so given studies show every $1 given to people on lower income results in around $1.2 in GDP* you'd expect overall GDP to grow provided its the UBI level and tax system is properly balanced.

Clever use of UBI also allows redoing many labour laws & protections because having UBI to fall back on will substantially improve the bargaining power of workers so they won't need the same protections.

* Sorry don't have a reference, seen this discussed a few places as common knowledge.

Just to add here that the columns and not just the rows are also sampled at each node so the trees are deliberately prevented from learning too much. This helps improve diversity and reduce overfitting. This is typically around 1/3rd of the number of columns and controlled by the mtry parameter.

Its an interesting question; if you're an aggressor and see a 'low tech' civilisation recklessly build an obvious Dyson sphere what probability do you put that its a honeypot from a real threat... and so on.

I personally find the dark forest interesting not because I think it is likely but because even giving it a very small probability of being true would be enough for near immortal high tech civilisations to move very cautiously.

Have there been any studies about whether Dyson swarms harvesting under 0.1% of the star's output are detectable, especially if the designers want to minimize the chance of detection?

I suspect that civilisations that make it to the Dyson swarm stage are on average a little biased towards environmentalism in order to survive past their terrestrial industrial stage.

Doesn't make sense to put a giant target on yourself until you can discount the dark forest hypothesis either, especially not when you've just beaten incredible odds to become a near-immortal truly space-faring civilisation.

As a gross approximation, the hormones and other chemicals might act like hyperparameters in the neural networks. Increase learning rates, adjust regularisation penalties, few-shot learning vs generalisation etc.

I'm kind of surprised no insurance company has started offering medical tourism yet... go to US hospital for this treatment or get a holiday and treatment in one of these destinations.

(edit: I'm not in the US so no idea what is typical outside the stories I see in the media)

This is good advice, will read the article now. Just checked out lofi, looks really great actually! Been looking for a minimalistic spotify visualiser / interface for my linux htpc.

I was thinking more ram, faster ssd or better gpu... but a browser accelerator that worked would be pretty cool :)

Its not unreasonable, people frequently talk about hundreds or even thousands of tabs open here on HN.

I have 8 virtual desktops with different projects open, R studio and Jupyter notebook windows plus reference reading material. The work I do typically involves long running tasks and checking back on projects or doing some add on analysis. It takes way too long to get everything aligned right and started up on different monitors so this machine is always on.

I'd expect web browsers to be able to handle this without ui lag if the hardware is good enough. At this rate, I might run firefox in separate docker containers per virtual desktop just to keep it smooth. Would benefit security too...

I would _love_ if firefox handled lots of background tabs better, my biggest complaint is how the browser UI lags with lots of background tabs open. This seems like its a scheduling issue, operations in background tabs should be lower priority than foreground tabs and should be pre-emptable so the user interface never lags. I'd be ok with manually pinning tabs I needed to run javascript in the background.

This is on a high spec dual 8 core xeon workstation with 164gb of ram and a radeon rx480 running latest kubuntu. Less than 100 tabs open, most of them basic references. If there was a bit of hardware I could buy that would fix browser lag I would.

DNN's make good features though, especially if you have time series data or lots of text.

I agree that the final model should be a randomforest/xgboost/lightgbm for typical tabular data.

If your concern is that it will reduce the supply of blood to younger people who currently use it then this is something that can be solved with incentives in a way that everyone benefits (e.g. people using it for rejuvenation pay more to subsidise higher production and are a lower priority when there is insufficient supply in the emergency room etc).

If on the other hand, that is just an excuse because you're morally opposed to old rich people living longer... ?

I'm honestly pretty conflicted here and the best thing that is coming out of this is some healthy debate.

IMHO the strongest argument against publishing it is because its built from Reddit it includes a lot of toxicity... VinayPrabhu originally made that argument here https://medium.com/@VinayPrabhu/reddit-a-one-word-reason-why...

Can't blame them for not opening up that can of PR worms. It could be like a version of Microsoft Tay that can form coherent paragraphs.

The bigger picture is this could be used for some very effective social media manipulation. A bit of clever infrastructure, maybe another generation of improvement and some understanding of human psychology (overton window etc) and managing 100,000 believable human accounts on something like reddit that subtly move debate becomes very doable.

Then again its going to happen anyway so better focus on the defensive techniques now...

Would be cool if this can be integrated with mpv, I tried SVP but integration is a bit awkward when you just want to chill.

That said SVP is nice for 30 fps youtube videos on slower connections or just videos without a high-quality 60hz source... even without svp videos look significantly nicer in mpv than in chrome on youtube.

This is a great resource, thanks for sharing!

I'd be interested to hear what kind of experience people are having with these frameworks in production.

Will scikit learn and pandas work without issue as well? Keen to try PyPy once I can run a typical data science stack.

Arguably computer graphics are the best candidate since many engines already have a velocity buffer for effects like temporal anti-aliasing and motion blur... would need API integration though.

My hunch is motion interpolation is preferred at high fps. E.g. a high action scene at 30 fps will need a lot more information to be filled in and has likely already been mastered with the target frame rate in mind (camera shutter speed etc).

But 120hz or more content will have very little motion between frames, and this motion will be far more linear (e.g. a finger might move 1cm between frames, but doesn't bend so less motion artefacts).

So 120hz content upscaled to 960hz+ would be much closer to native 960hz content and preferable to displaying at the original 120hz. A logical implementation would be VR headsets with a built-in DSP.

A bit off topic but I've been thinking there has to be a better way of handling this at the LCD controller level here...

Currently, a 60hz screen isn't perfect and pixel overdrive will be used to reduce motion blur. In effect, there are more 'unique intermediate frames' being displayed as the system tries to approximate the discrete input frames physically. The actual blur on fast-moving content on a modern screen is huge...

Ignoring latency for a moment, if a controller has the previous and next frame plus motion vectors and knowledge of the precise pixel response and decay curves for the monitor... it should be possible to continuously adjust the current/voltages(?) to every pixel to better approximate what a much higher refresh rate low motion blur panel would look like on high fps content. Can't remove blur but can make motion continuous rather than a stair step. A kind of smart motion interpolation + overdrive implementation.

The assumption is the higher fps the more linear the motion is frame to frame. 30 to 60fps interpolation should have more noticeable abnormalities than 120 to 1000hz.

There is a lot of data science work even in the low thousands of samples. Arguably most of the work will be on medium size data.

Technically every hypothesis that fits our data should get a little weight, plus our best available models of the universe show there should be billions of planets as habitable as ours.

That is evidence in the abstract sense and we currently lack the technology to observe these candidate planets in enough detail to test the models further.

We're like a person sitting in the desert at night with nothing but a candle concluding that the planet is completely deserted because they can't see anyone else.