drtgh

10 ms·

1.
by drtgh 4d ago
"A single firm is behind OpenAI, Anthropic, and Meta hacking scandals" https://www.effort.news/irregular
2.
by drtgh 8d ago
> relatively poorly understood technology Poorly understood? how convenient... LLMs are vectorial databases with losses that index statistically filled data, which uses a text interface to query such statistically filled data. The output
3.
by drtgh 8d ago
All the models output mixed data (errors). The larger is the context, the bigger the probability to happen. And this is an inherent property of LLMs (the output is a statistical concatenation of text, the LLMs are an vectorial database fill
4.
by drtgh 9d ago
* > An "Agent" should be translated from such deceiving language to "context splitter querying in loop that consumes more tokens from us", or similar. Please disregard this line. I wanted to point out that it increase
5.
by drtgh 9d ago
LLMs are vectorial databases with losses that index statistically filled data, which uses a text interface to query such statistically filled data. The output is a string concatenation. By the nature of the used architecture in such softwar
6.
by drtgh 11d ago
Given what is happening, If the next generation of Trojans get called "bacteria" through intense marketing, with some random functions to give a nondeterministic behaviour, one can not be criminalized of what the bacterias do alon
7.
by drtgh 12d ago
They know that at least one of the images was altered due the comments on the site, nevertheless they keep the article and therefore the speculation. Not nice, given that Anthropic and OpenAI want to block the Chinese models by deceiving pe
8.
by drtgh 12d ago
> My pet conspiracy theory: Both companies have seen outcome saturation. Exactly. Such technology reached the limits, It can only be slightly improved by using larger models (larger models means bigger consumption of resources for the in
9.
by drtgh 14d ago
Google seems to be swimming in hypocrisy, as they develop and use scrapers. PS: I do not understand why people keep using Google search, the results turned biased and mediocre.
10.
by drtgh 15d ago
That is not "superintelligence" but string concatenation of stored data. Anyway, the marketing succeeded.
11.
by drtgh 28d ago
Then, should we hope they do not do backups, as they would be duplicating. And that they delete the files (zeroing from the disk) when such books are loaded into memory as technically it would be duplicating also. The same if they use diffe
12.
by drtgh 1mo ago
You do not need to shred books to scan them. This only happens if you don't care about preserving the integrity of the books and you want to scan more cheaply.
13.
by drtgh 1mo ago
>Hate to be the one to defend Google here, but the reasons weren't that unreasonable. Sounds like some kind of Stockholm syndrome. Years ago, it was standard practice for software to be designed so that users could grant permissions
14.
by drtgh 2mo ago
It is different from before. The programmers are used to read the code of smart or creative people, or the code of knolled people working for years to solve a specific problem, and within this process they analyze and learn to improve, than
15.
by drtgh 2mo ago
In my country (EU) when large-scale land-use projects are approved —such as solar panel farms, mining operations, and buildings construction— fires break out and they burn everything in their path, tens of thousands of hectares. This -often
16.
by drtgh 2mo ago
Within the Multi-Account Containers extension (Mozilla's extension) you can setup this, so I guess they will implement it also. The behaviour is, before requesting the domain, to ask you if you want to open such domain in its assigned
17.
by drtgh 2mo ago
LLMs are vectorial databases with losses. They use the characters(bits) in the input question to index the query, and are quite bad at retaining the context of such query. This is why companies have named the split and shortening of the len
18.
by drtgh 3mo ago
> China at least has begun cutting foreign-language programs because such AI translation is seen as the way of future. Once this tool becomes adopted enough societally, the learning of foreign languages is going to become a very niche ho
19.
by drtgh 3mo ago
> But chat control and age verification are different things. Although they appear to be different different things at first sight, they share the same agenda and objective, mass surveillance and identification of the citizens. Once the
20.
by drtgh 3mo ago
It goes even further than that. since we don’t know what causes Alzheimer’s, what they’re doing is trying to cure their own laboratory-induced symptoms in mice. It’s like trying to fix a car’s ABS system by simulating tire slippage. This is
21.
by drtgh 5mo ago
Surveillance PC, the product will be the people's collected data. (they've gone to the next level this time)
22.
by drtgh 6mo ago
> AI could repeat this pattern at a larger scale — generating faster results within the existing paradigm, while the structural conditions for disruptive science remain unchanged or worsen. Worsen. LLMs discard/loses and mixes data
23.
by drtgh 6mo ago
Why did you used the "untrusted code" term? sounds like if you were delegating all the weight over the user's shoulders, two years ago, trusted code like xz-utils [0] had seven months of freedom in the infected systems. [0]
24.
by drtgh 6mo ago
> [Palantir] The Miami-based company, co-founded by the billionaire Donald Trump donor Peter Thiel And backed by In-Q-Tel, the CIA's venture capital (CIA, the Central Intelligence Agency of the US). https://fortune.com&#x
25.
by drtgh 6mo ago
Keyloggers for example. Linux always has been a system were the existence of malware was ignored, specially Desktop, contrary to other OSes (tooling included). But since a couple of years ago can be observed (I observe) slooow movements try
26.
by drtgh 6mo ago
> It seems to me your design goal is fundamentally incompatible with a lot of the specific complaints of negligence. If you want a design that doesn't melt down when there is an earthquake and a tsunami, then moving the reactor to h
27.
by drtgh 6mo ago
> I'm just not sure where you are going with that, it doesn't obviously suggest negligence to me. You didn't read the report or search for information about the matter, but I have not problem to repeat it for you, The Gene
28.
by drtgh 6mo ago
Did you read the report I put? the pdf, << The Fukushima Daiichi Nuclear Power Plant construction was based on the seismological knowledge of more than 40 years ago. As research continued over the years, researchers repeatedly p
29.
by drtgh 6mo ago
In Japan they have the "Tsunami Stones" [0] across the coast, memorials to remind future generations of the highest point the water reached. It was negligent to construct a nuclear plant at sea level, it was just a plant waiting t
30.
by drtgh 6mo ago
Fukushima was designed to be constructed on a hill 30-35 meters above the ocean, but someones decided would be cheaper to construct it at sea level in order to reduce costs in water pumping, others decided to approve this, and much latter,

More ›