cosmojg

A bar in my Italian family's hometown sells beers for one or two Euros a pop. The gelateria down the street charges a bit more for ice cream, though.

Ooh, this is great! Reminds me of one of my all-time favorite open-source Android apps: Sky Map! https://github.com/sky-map-team/stardroid

I think it's more likely that the big studios will start rolling out trivial offline modes (less risky) rather than overhaul their revenue models (more risky).

This is true of all markets.

They still come in real handy when you're stuck hacking away in a windowless SCIF.

How are you running the base model?

Look up whatever regional burns[1] exist in your country or neighboring countries, and attend one of those. Ideally, join whatever online community exists around it first, and then reach out to some of the camps that interest you about joining and helping with whatever it is they like to do. I love Burning Man, but I can honestly say that I've had a lot more fun at my local regional burns than I've had at the big burn.

[1] https://burnerguides.com/europe-burner-events

That's not a bad idea. It actually sounds like it could be a very useful hedge/insurance play.

This is certainly bad news, but at least an escape hatch exists (the "advanced flow") and it appears to be a one-time pain in the ass. If that changes, I hope GrapheneOS and friends[1] can get Google Pay or some alternative working so I can comfortably jump ship, as I rely pretty heavily on the ability to pay with my phone.

[1] https://eylenburg.github.io/android_comparison.htm

Oh wow, yeah, this is clearly a parody of that IPv8 draft. Just look at the ASCII tables! I suppose it's a commentary on how low the bar is for submitting drafts to the IETF.

Also, isn't insider trading the whole point of prediction markets? They were originally conceptualized as information aggregators.

This is mine: https://cosmo.red

Inspired by: http://bettermotherfuckingwebsite.com and https://github.com/morhetz/gruvbox

How comfortable is it? Do you have to wear it particularly tightly around your wrist or can you let it hang a little loose like a watch?

No, the Internet Archive is an organization that runs a web archive at archive.org while archive.is is an alternative domain for archive.today, a competing web archive run by an individual.

I must admit, one of my favorite recent-ish Android[1] features is that all text is made selectable in the app switcher using on-device OCR. Regardless of the app[2], you can just swipe up and start selecting text.

[1] ...at least on the Google Pixel.

[2] ...unless it's a banking app and it blocks permissions for screenshots and similar things.

Have you tried Google's Gemma-3n-E4B-IT in their AI Edge Gallery app? It's the first local model that's really blown me away with its power-to-speed ratio on a mobile device.

See: https://github.com/google-ai-edge/gallery/releases/tag/1.0.3

The voice sounds great! I find it quite aesthetically pleasing, but it's far from genderless.

Host an event! My friends and I host a "project night" every Thursday where we invite people to bring whatever personal projects[1] they're working on, and we strongly encourage plus-ones. As long as you have a venue (we started in a bar about 45 minutes west of Boston) and know at least one or two other people who are willing to consistently show up every week and you all put the work into inviting new people that you like and encouraging the people you like to invite new people that they like, the momentum will build on itself such that you'll inevitably find yourself with a thriving community, undoubtedly meeting at least one or two totally new people (on average) every week!

[1] Projects have included: crocheting a bunch of adorable little stuffed cows, developing a privacy-first webapp for visualizing social graphs, painting the cosmos with oil on a wall-sized canvas, engineering the control system for a human-scale vending machine, building an automatic water turret which uses image recognition to shoot squirrels with water when they climb on a bird feeder, planning and beta testing LARPs, board games, social experiments, and so on and so forth.

To my eyes, it looked like TV snow made up of blue and green particles! Twas quite delightful.

For what it's worth, I think Kimi's modified MIT license still meets the OSI definition of "open source." For example, the explicitly OSI-approved "Attribute Assurance License"[1] contains similar wording:

each time the resulting executable program or a program dependent thereon is launched, a prominent display (e.g., splash screen or banner text) of the Author’s attribution information

[1] https://opensource.org/license/attribution-php

Oh man, Brut's HIPPOCRATIC LICENSE[1] is interesting! Just excerpting bits of Section 3.1 and Section 3.2 for example:

  * 3.1. The Licensee SHALL NOT, whether directly or indirectly, through agents
   or assigns:

    * 3.1.12. Taliban: Be an individual or entity that:
      
      * 3.1.12.1. engages in any commercial transactions with the Taliban; or
      
      * 3.1.12.2. is a representative, agent, affiliate, successor, attorney, or
        assign of the Taliban;
   
    * 3.1.13. Myanmar: Be an individual or entity that:
      
      * 3.1.13.1. engages in any commercial transactions with the
        Myanmar/Burmese military junta; or
      
      * 3.1.13.2. is a representative, agent, affiliate, successor, attorney, or
        assign of the Myanmar/Burmese government;
   
    * 3.1.14. Xinjiang Uygur Autonomous Region: Be an individual or entity, or a
      representative, agent, affiliate, successor, attorney, or assign of any
      individual or entity, that does business in, purchases goods from, or
      otherwise benefits from goods produced in the Xinjiang Uygur Autonomous
      Region of China;

  * 3.2. The Licensee SHALL:

   * 3.2.1. Provide equal pay for equal work where the performance of such work
     requires equal skill, effort, and responsibility, and which are performed
     under similar working conditions, except where such payment is made
     pursuant to:
     
     * 3.2.1.1. A seniority system;
     
     * 3.2.1.2. A merit system;
     
     * 3.2.1.3. A system which measures earnings by quantity or quality of
       production; or
     
     * 3.2.1.4. A differential based on any other factor other than sex, gender,
       sexual orientation, race, ethnicity, nationality, religion, caste, age,
       medical disability or impairment, and/or any other like circumstances
       (See 29 U.S.C.A. § 206(d)(1); Article 23, United Nations Universal
       Declaration of Human Rights; Article 7, International Covenant on
       Economic, Social and Cultural Rights; Article 26, International Covenant
       on Civil and Political Rights); and
   
   * 3.2.2. Allow for reasonable limitation of working hours and periodic
     holidays with pay (See Article 24, United Nations Universal Declaration of
     Human Rights; Article 7, International Covenant on Economic, Social and
     Cultural Rights).

You can certainly take notes on all that anecdata and use them to design some proper experiments! But until someone gets the resources and willpower together to do all of that, the anecdata is all the evidence we have.

Both the Rationalist community and the Rust community are very active in pursuing their goals, and unfortunately, it's far easier to criticize others for doing things than it is to actually do things yourself. Worse yet, if you are not yourself actively doing things, you are far more likely to experience fear when other people are actively doing things as there is always some nonzero chance that they will do things counter to your own goals, forcing you to actively do something lest you fall behind. Alas, people often respond to fear with hatred, especially given the benefit of physical isolation and dissociation from humanity offered by the Internet, and I think that's what you're seeing here on Hacker News.

Sounds like they might be switching to a source-available model rather than an open-source model.

They actually use distilled versions. The most egregious example of this is their misleading reference to all distillations of DeepSeek-R1, which are based on a variety of vastly different base models of varying sizes, as alternative versions of DeepSeek-R1 itself. To this day, many users maintain the mistaken impression that DeepSeek-R1 is overhyped and doesn't perform as well as claimed by those who have been using the actual model with 685B parameters.

What do you use instead of a real phone number? How do you get past mandatory phone number verification for services which require it?

Having worked on a similar endeavor, I doubt they intentionally dragged their feet on it. They likely had a smorgasbord of legal bullshit and technical challenges resulting from code omissions mandated by said legal bullshit that they had to muddle through.

They are related by Landauer's Principle: E=k×T×ln(2), where E is the energy required to erase 1 bit of information, k is the Boltzmann constant, and T is the temperature of the system storing that 1 bit of information.

Information is physical. It is inextricably tied to the physical degrees of freedom of the system storing it. Per Landauer's Principle, erasing information is an irreversible process that increases the entropy of the environment, and this increase in entropy is the dissipation of energy. With that in mind, I would argue that you are correct, energy and information are in fact two sides of the same coin.

Uh, this isn't true at all. Did you actually look at the repo[1]? Only the metadata (i.e., LLM-generated task names and interaction classifications) have been made available.

[1] https://huggingface.co/datasets/Anthropic/EconomicIndex

Who in their right mind is going to blindly take the code output by a large language model and toss it on a cruise missile? Sleeper agents are trivially circumvented by even a modicum of human oversight.