d_silin

3 ms·

31.
by d_silin 2mo ago
Article gave me an idea that Indus Valley seals were baker's stamps - that's why the imprinted goods were never found (eaten!).
32.
Dartmouth Workshop on Artificial Intelligence (1956)(en.wikipedia.org)
2 points by d_silin 2mo ago|0 comments
33.
by d_silin 3mo ago
My radical opinion is that LLMs are harmful for software development - they are the ultimate "goto" operator. All actual code should be written by a human developer. Instead, use them in adversarial mode - run QA scenarios using L
34.
by d_silin 3mo ago
You can absolutely select founding population to be maximally genetically diverse and of large enough number to avoid inbreeding. Speciation is a likely outcome for interplanetary and interstellar colonies, yes.
35.
by d_silin 3mo ago
beats rawdogging it, at least.
36.
by d_silin 3mo ago
"The "S" in AI stands for security" strikes again. Run any cloud-based AI agents in VM/container and map your host's local folders to guest OS as needed. Takes more effort that default way, I know.
37.
by d_silin 3mo ago
One of the most interesting technologies that is not about LLMs/AIs.
38.
by d_silin 3mo ago
My bad! Global aerospace industry then.
39.
by d_silin 3mo ago
1969 was truly the pinnacle of US aerospace industry - Concord, Boeing 747 and Apollo 11 all happened during this year.
40.
Small satellite operators confront a bottleneck to space access(spacenews.com)
9 points by d_silin 3mo ago|0 comments
41.
by d_silin 3mo ago
This should be way higher.
42.
by d_silin 4mo ago
There is no metric a reasonably intelligent person can't sabotage or subvert... ...for example, you can write a script to burn tokens and write the code yourself.
43.
by d_silin 4mo ago
Nobody forces you to use AI. It has become sort of junk food for the brain. Temptations and ads for it everywhere.
44.
by d_silin 4mo ago
That's one idea I am willing to bet on (with my money and skills).
45.
by d_silin 4mo ago
Fair enough. I appreciate your civility.
46.
by d_silin 4mo ago
When does it stop, though? A white lie during interview, a deadline you promised and knew you could not deliver? A product you ship that claims to do X, but it doesn't?
47.
by d_silin 4mo ago
Culture begins at the front door, corporate or personal. I would expect absolute sincerity from pilot or a doctor during the interview, including history of mental health and professional mistakes. Authority over lives of people must come w
48.
by d_silin 4mo ago
Goes both way, yes. History has examples of extremely underprivileged people not compromising on the values even when facing death or torture.
49.
by d_silin 4mo ago
Having any kind of integrity is expensive, financially, emotionally and sometimes physically. Software development is not that high-stakes of a job anyway. There is always another interview. I got another one soon enough, where the employee
50.
by d_silin 4mo ago
Correct. I did it anyway (and yes, it was awful).
51.
by d_silin 4mo ago
In the end, everyone makes their own choices. I don't want to preach martyrdom, but I am also offended by people choosing moral bankruptcy when faced with even the slightest hardships.
52.
by d_silin 4mo ago
Yes.
53.
by d_silin 4mo ago
It is not. I made that choice in the past and will do it again. "Speak the truth, even if your voice shakes"
54.
by d_silin 4mo ago
The only effective action is push back on Athropic.
55.
by d_silin 4mo ago
SpaceX already blown through a number of "... but this cannot be!" forecasts. Would not discount them easily.
56.
by d_silin 4mo ago
TL;DR: - AI part of SpaceX is absolute disaster, while its IPO valuation hinges on its future success. - Lot's of questionable corporate shenanigans and accounting tricks. - Public shareholders have zero power over Elon after IPO. Some
57.
by d_silin 4mo ago
Entire civilization flourished for 2000 years and then disappeared without any clue why. I have a pet theory about Indus Valley script - inscriptions on the seals are so short and unique because they are just name signatures, to stamp other
58.
by d_silin 4mo ago
"Recursive self-improvement" is in the same league as "perpetual motion". What would be a way to recursively self-improve algorithms for matrix multiplication (foundations of machine learning and inference)?
59.
by d_silin 4mo ago
Who says "valuation"? I am talking about "profits", something that none of the the upcoming IPO debutantes have managed to achieve.
60.
by d_silin 4mo ago
Among IT giants, Apple is the only such option.

More ›