> It seems like developers, of all people, are willingly letting their code be AI piggybacked.
Is Codeberg actually effective at preventing crawling of public code they host?
2,070 karma · joined October 30, 2013
Is Codeberg actually effective at preventing crawling of public code they host?
* Rankings: https://data.worldhappiness.report/table
* Data for the US specifically: https://data.worldhappiness.report/country/USA
The main thing I found interesting was the big jump in "Benevolence" for 2020-2021, presumably related to the pandemic.
* some of the US government officials protected by the Secret Service were the targets of swatting
* the USSS found the swatting calls were anonymized by a SIM Farm in/near NYC
* their investigation of the SIM Farm found "300 co-located SIM servers and 100,000 SIM cards across multiple sites"
* it could have hypothetically been used for swatting officials at the UN General Assembly, but that seems to be conjecture by the Secret Service, rather than anything they actually have evidence of
Does that seem consistent with what we know?
> Fc-optimized CD40 agonistic antibody elicits tertiary lymphoid structure formation and systemic antitumor immunity in metastatic cancer
> CD40 agonism enhances antitumor immunity but is limited by systemic toxicity and poor efficacy. Here, we present a phase 1 study (NCT04059588) of intratumoral (i.t.) 2141-V11, an Fc-engineered anti-CD40 agonistic antibody with enhanced binding to the inhibitory receptor FcγRIIB. Among 12 metastatic cancer patients, 2141-V11 was well tolerated without dose-limiting toxicities. Six patients experienced tumor reduction, including two complete responses in melanoma and breast cancer. 2141-V11 induced regression in injected and non-injected lesions, correlating with systemic CD8+ T cell activation and mature tertiary lymphoid structures (TLSs) in complete responders. In CD40/FcγRs humanized mice bearing orthotopic tumors, i.t. 2141-V11 promoted de novo TLS formation, facilitating i.t. CD8+ T cell effector responses independent of lymph node priming. The resulting local immune responses by 2141-V11 mediated abscopal antitumor effects and sustained immune memory. These findings demonstrate that i.t. 2141-V11 is safe and promotes immune-privileged tumor microenvironments that promote systemic and durable antitumor immunity.
> A Federal District Court found that these raids were part of a pattern of conduct by the Government that likely violated the Fourth Amendment. Based on the evidence before it, the court held that the Government was stopping individuals based solely on four factors: (1) their apparent race or ethnicity; (2) whether they spoke Spanish or English with an accent; (3) the type of location at which they were found (such as a car wash or bus stop); and (4) the type of job they appeared to work. Concluding that stops based on these four factors alone, even when taken together, could not satisfy the Fourth Amendment’s requirement of reasonable suspicion, the District Court temporarily enjoined the Government from continuing its pattern of unlawful mass arrests while it considered whether longer-term relief was appropriate.
> Instead of allowing the District Court to consider these troubling allegations in the normal course, a majority of this Court decides to take the once-extraordinary step of staying the District Court’s order. That decision is yet another grave misuse of our emergency docket. We should not have to live in a country where the Government can seize anyone who looks Latino, speaks Spanish, and appears to work a low wage job. Rather than stand idly by while our constitutional freedoms are lost, I dissent.
> Weldon, a Florida representative from 1995 until 2009, has been critical of the federal government’s oversight of vaccine safety. While in the House, he sponsored a bill that would have taken vaccine safety oversight from the CDC and created a separate “Agency for Vaccine Safety Evaluation” under HHS, and as recently as 2019 repeated the disproved claim that “some children can get an autism spectrum disorder from a vaccine” in an interview.
> Place another AGENTS.md inside each package. Agents automatically read the nearest file in the directory tree, so the closest one takes precedence and every subproject can ship tailored instructions. For example, at time of writing the main OpenAI repo has 88 AGENTS.md files.
Funding info is at the bottom of the article, the project was primarily funded by the Swedish government.
* It focuses on "industrial emissions", which in their analysis doesn't seem to include agriculture
* It's largely indirect emissions, e.g. if someone uses gas in their car from a gas station which was refined from crude oil purchased from Saudi Aramco.
> Set your own rules: Customize Cursor's work with rules, AGENTS.md, and MCP.
There's no mention of it in the docs, though. It's also interesting it's AGENTS.md on that page instead of AGENT.md, I wonder if that's a typo.
I guess having links to supplementary rules files is an option, but I'm not sure which agents (if any) would work well with that.
Anecdotally, the tool updates in the latest Cursor (1.4) seem to have made tool usage in models like Gemini much more reliable. Previously it would struggle to make simple file edits, but now the edits work pretty much every time.
Your own experience isn't generally applicable. Although it was a couple decades ago, I still use various things I learned from my non-major classes pretty much daily.
The GitHub repos for Claude Code and Cursor are basically just issue trackers and some examples. I'm actually surprised they have as many stars as they do.