divbzero

How far away PLATO will be from the James Webb Space Telescope? How big is the L2 Lagrange point? (i.e., how closely do you need to be for an orbit around L2 to be practical?)

Web: https://bookshop.org/ebooks

Android: https://play.google.com/store/apps/details?id=org.bookshop.a...

iOS: https://apps.apple.com/us/app/bookshop-org/id6465290096

Yes, different training data could have an effect too, but that doesn’t appear to be what’s happening here.

I’ve seen examples of this being less of an issue when running DeepSeek locally.

With more efficiency offset by more light bulbs, electricity used for lighting has been roughly flat since 2010: https://www.iea.org/data-and-statistics/charts/global-electr...

(For LLMs I wish that efficiency could lead to less electricity used for chips, but I think the best we can hope for is for electricity use to flatten out.)

Greater efficiency of light bulbs has led to more light bulb use, not less. More efficient training of LLMs could just as likely lead to more chip use, not less.

… wings.

Where is the good old…

  <center>alignment with no CSS</center>

Disappointed to see <blink> or <marquee> merely mentioned but not used on that page, seemed like a serious omission.

Based on my tests just now, <blink> no longer blinks in today’s browsers but <marquee> still scrolls happily.

The article mentions that the galaxy is a big target for Hubble to image but doesn’t specify exactly how large: Andromeda stretches 3° across our sky compared to our Moon’s apparent diameter of 0.5°. It would be quite a sight to behold if it were bright enough to see by naked eye.

And yet, Andromeda is the closest major galaxy to the Milky Way: next door neighbors in the small patch of the Universe known as the Local Group.

As far as I can tell, the descriptions in your first and second paragraphs are in stark conflict.

The outgoing Biden administration actually stated that they wouldn’t enforce the ban for just one day, choosing to leave implementation of the law to the incoming Trump administration.

Efforts to save TikTok have been bipartisan (“Senate Democratic Leader Chuck Schumer said he spoke with Biden on Thursday to advocate for extending the deadline to ban TikTok.”) and efforts to enforce the ban have also been bipartisan (“Democrats had tried on Wednesday to pass legislation that would have extended the deadline, but Republican Sen. Tom Cotton of Arkansas blocked it. Cotton, chair of the Senate Intelligence Committee, said that TikTok has had ample time to find a buyer.”)

https://www.pbs.org/newshour/politics/biden-wont-enforce-tik...

Note that the outgoing Biden administration has stated that it won’t enforce the TikTok ban today, choosing to leave implementation of the law to the incoming Trump administration. TikTok shutting down today is an internal choice.

https://www.pbs.org/newshour/politics/biden-wont-enforce-tik...

It worked for me on the second try. The lines had to connect just right.

Empty set ∅ and therefore ∴ worked for me too.

Do you have links to the Human Rights Watch reporting that you reference?

> we will see more focused efforts to create truly free generative AI systems, perhaps including the creation of one or more foundations to support the creation of the models

What are the biggest barriers to making this a reality? The training data or the processing power?

Which open-source projects, if any, are the farthest along in this effort?

Previous discussions:

Oh Shit, Git - https://news.ycombinator.com/item?id=31874308 - June 2022 (232 comments)

Oh Shit Git? - https://news.ycombinator.com/item?id=24173238 - Aug 2020 (156 comments)

Oh shit, git (2016) - https://news.ycombinator.com/item?id=19906972 - May 2019 (277 comments)

Oh shit, git: Getting myself out of bad situations - https://news.ycombinator.com/item?id=15951825 - Dec 2017 (509 comments)

This is old school. No switch or restore here.

The article is focused on Nvidia, but note that Apple [1][2] and Google [3] have also been working in this area and will undoubtedly continue to do so.

[1]: https://developer.apple.com/machine-learning/core-ml/

[2]: https://machinelearning.apple.com/research/neural-engine-tra...

[3]: https://research.google/blog/improved-on-device-ml-on-pixel-...

Not everyone has low latency high bandwidth reliable internet, and almost no one has it all the time. In addition, the privacy and security benefits from running things locally is a requirement for some.

So I should really be booking from a logged out incognito browser on my Linux machine with VPN to a less wealthy region.

To expand on the headline: US employee engagement in 2024 was 31% which was tied with the lowest in the past decade (since 2014) but also higher than at any point in the preceding decade (2000 – 2013). Lowest engagement was 26% in 2000 and 2005. Peak engagement was 36% in 2020.

(Sample size was ~80,000 working adults. Sampling error was ±0.5 percentage points at a 95% confidence level.)

I was half expecting the outcome to be worse than it turned out. Two died while helping with the rescue, but all of the passenger and crew on the train survived.

She absolutely got it done, but not surprisingly there was still significant legal backlash she had to deal with.

https://www.nature.com/articles/nature.2017.22196

Doesn’t the small size make it more impressive that the Texians held out for so long?

getYear() must have been written in the 1990s during the early years of Java:

   95 = 1995
   96 = 1996
   97 = 1997
   98 = 1998
   99 = 1999
  100 = 2000 !?

I believe the layer you’re thinking of is ‘glaze’:

https://en.wikipedia.org/wiki/Ceramic_glaze

If I understand this right, Google isn’t actually selling user data to auction losers because auction losers are collecting the data without spending any money.

> Fengning’s advanced design includes an upper reservoir with a capacity of 45.04 million cubic meters and a lower reservoir holding 71.56 million cubic meters.

A cubic meter of water weighs a tonne. So that’s literally tonnes and tonnes of water.