You're conflating concerns. Most government systems are required to log the hell out of their inputs and outputs. Making decisions to destroy something based on ephemeral data could happen just as easily on the ground as it could in orbit -- it has nothing to do with what kind of system (large neural network, traditional ML, human decision, or something else) makes the decision or where the decision happens.
I'm not sure what you mean. Do your eyes count as "long range sensors"? Mine can see stars that are many light-years away, and eyes are not a new invention.
They claim to have realized (invented) a better quantum magnetometer and a way to process the data to do a pretty amazing kind of detection. That's one very specific kind of long range sensor, with improvements over previous long range sensors but also limitations of its own. It's presumably not a magic device that Pareto dominates other long range sensors.
What's the differential latency of running a strong model for several turns (or the equivalent) on a spacecraft's power budget compared to a data center's power budget, especially once you factor in redundancy to manage single-event upsets in the huge RAM array needed for that model?
I use Claude rather than a local model because I don't want to wait all afternoon for the quality of results I can fit into 128 GB RAM.
"away from densely populated areas" is both subjective and not the same as "away from other development".
This kind of article should wait until the study has been peer-reviewed, not flood the zone with unreviewed slop.
I understood the question to be whether the study controlled for other changes in land use in the surrounding area. For example, northern Virginia has built a ton of new data centers close together over the last decade -- in many cases, replacing pastures. Attributing the results of the whole set to individual data centers would be a methodological error.
They have working software: OCX is intended to replace the current ground system, Operational Control System (OCS). They have launched a lot more than 30 satellites -- in fact, most of them have been decommissioned, although the currently operational set have mostly outlived their design lifetimes by a lot. The oldest active satellites are Block IIR satellites, with a design life of 7.5 years
That doesn't at all answer the question that was asked.
Parents have extremely broad rights to manage their children's upbringings. Kids have no right to use end-to-end encryption without parental consent, although I don't think a court has held that parental consent is necessary.
Sure, it is not a big problem for SSH. It is a problem when you connect to a web site, especially as certificate lifetimes get shorter: you need the whole certificate chain from a root (that your browser trusts) to the web server, which means at least two public keys and signatures and often more.
The NIST-approved post-quantum options and PK/sig sizes (in bytes, for "security level 1", which is the lowest) are Crystals Dilithium 2 (1312 / 2420), Falcon-512 (897 / 666 but computationally expensive) or SPHINCS+-SHA2-128s (32 / 7856 for the smaller but more computationally expensive signatures; same for SPHINCS+-SHAKE-128s). This compares to 32/64 or 64/48 bytes for 256-bit ECC algorithms and 256/256 bytes for 2048-bit RSA. If you are fetching a few kilobytes of text or CSS, this additional overhead is huge.
Yup. I'm waiting for any quantum computer to actually break a non-trivial public key, even of a laughably small order (like RSA130, which was factored by classical computers 30 years ago). Lots of people get famous for papers based on theoretical quantum gates that nobody knows how to realize.
Elliptic curve crypto is vulnerable to the same kind of theoretical quantum attacks as integer-factorization cryptography. You currently need to use algorithms with unfortunate trade-off (large public keys or large signatures/key agreements) to get resistance to quantum attacks.
Assuming quantum computers ever factor numbers larger than 21 without cheating or falling back to deterministic algorithms, at least.
That, and "gainful employment" usually means a person has a job with pay that meets their living expenses plus a bit. What living expenses do these robots have, except for the livings they take from humans?
He's also a fool when it comes to politics. People should want government writing rules and picking winners for AI just as much as for social and news media: not picking winners at all, and setting as few and as narrow rules as possible. Do you want the default (or only) AI service to run like the DMV?
Betteridge's Law continues to hold true.
My new open source-as-a-service is going to be called Compu-Global-Hyperega-Clawed. Matt Growing and Anthropic are both cool with that, right?
Assuming a 2027 release: EntraBox Series CoPilot 2028 Community, Available In Azure.
It is not every question that deserves an answer. -- Publilius Syrus
