30,560 karma · joined December 4, 2018
Both Vyr Cossont and Ngaroe QiRia are so bored with their meaningless existence that they have to go out of their way to invent “life tasks” and other challenges to immerse themselves in.
Not to mention the scene where a ship mind casually crushes the spirits of Vyr by perfectly playing a piece on its first attempt that she struggled with for years!
I ignore every official statement about “support” and simply look at the horse's mouth: Microsoft hasn’t bothered to fix this in IIS, and likely never will.
It’s a dead product, “supported” only in a contractual sense, not a practical one.
(If you called support right now demanding fix for this, they’d tell you that it’s broken by design, have you tried moving to a recurring revenue service offering?)
Solar and wind have been branded as unreliable by propogandists working for billionaire fossil fuel magnates, oil sheiks, and oligarchs.
The converse is true: distributed power generation and storage is incredibly robust and can easily meet the needs of our civilization even under active attack.
If you want long term thinking, if you want robustness over short term profits, then solar + wind + batteries are you answer.
No super power can blockade the sun and wind.
No super power can blow up every rooftop panel without also nuking entire cities.
Humanoids in the Culture are controlled at every level, even their thoughts as guided… sorry… “aligned” with the Minds through the adoption of the language Marain.
No humanoid ever overrules a ship Mind in the novels.
(Several characters in the novels comment on this, so it is entirely cannon.)
The Culture fought a war against the Idirans, who eschew the use of intelligent machines. The Idirans lost.
Similarly, the Gzilt civilisation in the novel The Hydrogen Sonata prefer to run their ships with a quorum of uploaded humanoid mind states running at highly accelerated speeds.
GPS is entirely passive (it is receive only), so leaving it on is harmless for your privacy on Apple devices. For Android... well, if you have an Android phone, you're not a privacy-conscious consumer by definition, so just leave it on for the convenience of not having to turn it on and off manually all the time!
WiFi, again, is not that big a deal with Apple because they rotate your unique MAC address to preserve your privacy. Android is a mixed bag, depends on the model of phone you purchased.
It's the equivalent of making pipe bombs at home.
The instructions are out in the public, but that's not a problem because the idiots that make pipe bombs blow themselves up more often than they succeed in doing terrorism.
Dario keeps saying in interviews that AI hacking is not a problem because "the good guys will protect themselves with AI", but then gate-keeps.
I submitted the form and it was rejected without reason.
Now what?
Concepts cars are cars.
What was shown on stage is closer to the Apollo lander training rigs than a vehicle that you could purchase and drive on a public road.
It can never, ever be sold as a production vehicle! The COPV gas storage is under enormous pressure, and would explode with the force of a decent sized bomb if damaged in an accident.
These things would be regularly turning their owners into pink mist, emulating Stockton Rush at a larger scale.
I regularly run "experiments" on Azure to figure out what it actually does.
Out of curiosity, I just tried Gemini 4 Argon and it seems to have nailed it, decoding even your "illegible" words!
See: https://aistudio.google.com/app/prompts?state=%7B%22ids%22:%...
These modern languages have both easy-to-use "owning" string types and a "string slice" type that can be passed around efficiently. Unlike C, in these languages a string and a slice are largely interchangeable, which makes them easier to use throughout not just your own codebase, but third-party libraries too.
For example, the Rust `httparse` crate `Request` type returns parsed fields as string slices, not strings, which means it doesn't allocate anything on the heap: https://docs.rs/httparse/latest/httparse/struct.Request.html
Err... doesn't that completely invalidate the big-O analysis!?
I suspected that these Open AI "proofs" of sub-nlogn times were cheating somehow, by sweeping something under a rug. This seems to be it!
That five trillion dollar market cap has an expiry.
The “bottom end” of the camera market is basically gone now, but the high end full frame bodies are still selling.
The key points are that this was a lawsuit by a rival candidate, not Microsoft, and was overturned by a court of appeals because it was without merit: the candidate Mikhail Matveyev used the Arial and Tahoma fonts in a report linked from his campaign website, which the court deemed not to be part of his campaign materials. Not to mention that having a license for either Office or Windows will give grant the rights to use those two fonts!
In other words a rival tried and failed to get him on an invalid technicality.
That's easily fixed!
Spinning a space station for gravity is straightforward, it's just that the ISS wasn't designed for it, and NASA didn't have the budget to build a second, bigger space station with "extra engineering challenges" on top.
It's definitely practical though, it's just a matter of launching enough mass to orbit.
Like you said, Starship is going to change a lot of things, starting with space station design.
This is like being arrested for driving on the right in the United States.
Also, the tree can be based on any underlying hash function, so the root hash can be literally "just" a SHA256 hash, but instead of hashing the whole file, it hashes a list of hashes.
You don't even need to store those intermediate nodes! They must be recomputed on-the-fly to verify the root hash anyway.
Last, but not least, the temporary space required is logarithmic in the size of the data, typically with a large exponent, so in practice this all fits on the stack. No heap memory allocations required, compatible with embedded devices, and can re-use cryptographic acceleration primitives like the AES-NI instruction set.
Conversely, tree-based hash constructions allow:
- Parallel generation and verification.
- Incremental repair of partially corrupted data.
- Extension of an existing hash (efficient append & prepend).
- Patricia hashes also allow efficient mutation in the middle of the data.
- Various deduplication and efficient incremental sync scenarios.
America is lost.
Quality and veracity are independent (orthogonal) axes, but we are used to them being correlated, so some people conflate the two.
A knock-off Chanel handbag likely won’t be as good quality as the original, but it could be even better! Even so, it wouldn’t be authentic.
The leaves and then all nodes can be computed in parallel.
With a typical node size of say 4 KB any file that is at least 4 MB in size can utilise 1,024 cores.
Most of the DB engines out there still seem to use a "n-threads" style parallelism with exchange operations and poor async I/O management.
DuckDB is improving on this front, but in some sense is catching up to R&D (and implementation!) that is now decades old.
A "rhetorical challenge" I like to give software developers working on systems like this is the following: If I gave you a computer with 1,024 cores and matching network and storage bandwidth -- but with significant latency -- could you keep a system like this 100% utilised with one query?
The answer for almost all software is "no".
For example, SQL Server tops out at 64 hardware threads for any one query: https://learn.microsoft.com/en-us/sql/database-engine/config...
GPU codes are starting to get there, but CPU codes are way behind on this frontier of computer science.
It's not just databases! Can you (de)compress a file in parallel? Verify its hash in parallel? Upload/download from storage with CPU and I/O task parallelism? Can you overlap all of these operation so nothing is ever waiting on anything else it doesn't have to?
This matters! I ran some tests with bioinformatics codes and found that most got stuck in tar pits. Many could not scale to modern SSDs with millions of IOPS or modern networking with hundreds of gigabits of throughput, no matter how many CPU cores were thrown at them.
PS: AMD's Zen 6 era EPYC 9006 processors will have 512 cores and 1,024 threads per two-socket system, so this is not hypothetical: https://www.amd.com/en/products/processors/server/epyc/9006-...
In other words, they’re selling a product to their buyers, and the product is politicians.
That's literally what happened. Thousands of .si domains were bought up before his renaming, and then offered up for millions.
PS: His wife is from Slovenia. I'm sure that's a coincidence...
The "expensive part" of generating the next token is streaming in the model weights from memory. The computations are relatively simple, which is called a "low arithmetic intensity" in industry jargon.
So what they do is batch multiple chats together and compute the neuron activations for all of them together.
This is vaguely similar to how some database engines work, where if multiple users need to run a "whole table scan" query, the additional users "join" the streaming workload of the first query mid-way, then loop back around to complete the first part that they missed. The AI accelerators don't do this looping, but the concept is the same: amortize the expensive I/O over multiple computations running in parallel.
The "turbo mode" token rate thing is almost certainly your query getting sent to slower or faster hardware, like B200 vs newer B300 kit.