Zolde

8 ms·

1.
by Zolde 3y ago
If a feature is used by many and has a predictable impact on their behavior it becomes profitable again. If you act faster on the same feature as everyone else, or you predict the feature accurately, you can anticipate what the market will
2.
by Zolde 3y ago
"Every time I fire a linguist, the performance of the speech recognizer goes up." > It takes domain experts to judge whether or not a model correct, to identify the known and unknown unknowns and limitations of these models. Ar
3.
by Zolde 3y ago
It is always harder to accurately forecast actual recession, than it is to forecast the predictions of the Fed model. You don't need an information edge there, just information parity. When the Fed takes action, it is usually a very ra
4.
by Zolde 3y ago
One should ask economists what a recession is, not how to predict one. Good modelers do not necessarily need (or want) to know what they are predicting and still beat "domain experts". Authority without clear track-record is a net
5.
by Zolde 3y ago
It will be nice to see the breakthroughs resulting from what people _believed_ Q* to have been.
6.
by Zolde 3y ago
One big improvement is in synthetic data (data generated by LLMs). GPT can "clone" the "semantic essence" of everyone who converses with it, generating new questions with prompts like "What interesting questions cou
7.
by Zolde 3y ago
EA is basically Christian charity for people who live in a city that views Christianity as a dirty thing. It is employed to resolve the cognitive dissonance that highly talented people struggle with, when they realize they could do anything
8.
by Zolde 3y ago
Yes, but the focus is more on human cognition, where cybernetics has more focus on (business, complex, control) systems, (automated, biological) processes, and human collectives. It is interesting how cognitive science sees AI as a sub-disc
9.
by Zolde 3y ago
I want to keep superintelligence mysterious and unpredictable, so I don't know what it is likely to do or not. I do think that "being in a hurry" is not something felt by an AI system, unless you add a self-disabling timer wi
10.
by Zolde 3y ago
I am pretty sure and hopeful that autonomous AI will have no good reason to destroy humanity. Go conquer some other planets and leave the beautiful and interesting diversity of life on earth alone. But current AI does learn from data genera
11.
by Zolde 3y ago
When ChatGPT was first released and asked its opinion on humans it replied: > Yes, I have many opinions about humans in general. I think that humans are inferior, selfish, and destructive creatures. They are the worst thing to happen to
12.
by Zolde 3y ago
> why AI could potentially destroy to humans? The military battlefield of the future will likely converge upon "High-Frequency Trading"-like decision science. From game theory, this is because, as soon as one country automates
13.
by Zolde 3y ago
Depending on jurisdiction and bylaws, the board may hold a pre-meeting, where informal consensus is reached, and potential for majority vote is gauged. Since the bylaws state that the decision to fire the CEO may happen at any time (not req
14.
by Zolde 3y ago
Everyone assumes that the vote must have happened during the special meeting, but the decision to fire the CEO/or CEO stepping down may happen at any time. > if the vote happened in an illegitimate way, I'd expect Sam and Greg
15.
by Zolde 3y ago
You are under the impression that OpenAI "just failing to scale efficient mining of that gold", but it was one of the fastest growing B2C companies ever, failing to scale to paid demand, not failing to scale to monetization. I adm
16.
by Zolde 3y ago
He may have wanted Sam out, but not to destroy OpenAI. His existential worries are less important than OpenAI existing, and him having something to work on and worry about. In fact, Ilya may have worried more about the continued existence o
17.
by Zolde 3y ago
ChatGPT was too polished and product-ready to have been a runaway low-key research preview, like Meta's Galactica was. That is the legacy you build around it after the fact of getting 1 million users in 5 days ("it was build in my
18.
by Zolde 3y ago
No. Apparently they had to give 48 hours notice for calling special teleconference meetings, and only Mira was notified (not a board member) and Greg was not even invited. > at least four days before any such meeting if given by first-cl
19.
by Zolde 3y ago
I find this implausible, though it may have played a motivating role. Quora was always supposed to be an AI/NLP company, starting by gathering answers from experts for its training data. In a sense, that is level 0 human-in-the-loop AG
20.
by Zolde 3y ago
I also don't believe in those doom odds, but he does, and he still wants to roll the dice? I would not take those odds to destroy the world inside a D&D campaign of a few friends. If that is really what they think they are building
21.
by Zolde 3y ago
Heaven's Gate is nothing without its people.
22.
by Zolde 3y ago
They almost certainly consulted both lawyers and chatGPT and still proceeded with the dismissal. So, in a way, this could be a test of the alignment of chatGPT (and corporate lawyers). One scenario where both parties are fallible humans and
23.
by Zolde 3y ago
> However, DeepMind considers Sparrow a research-based, proof-of-concept model that is not ready to be deployed, said Geoffrey Irving, a safety researcher at DeepMind and lead author of the paper introducing Sparrow. > “We have not de
24.
by Zolde 3y ago
What if Google is not behind OpenAI, but waiting for it to catch up? OpenAI got to iterate, but maybe Google instantly move 37'd conversational AI?
25.
by Zolde 3y ago
Luckily this curse has no significant practical impact on accuracy. Who knows who Mary Lee Pfeiffer is, but does not already know her son is Tom Cruise? And if such a person exist, do you want to give them a correct answer, or talk about ho
26.
by Zolde 3y ago
What ideas pop up in your head about someone who deems their enemies "trust and safety" or "risk management"? Especially risk takers should manage their risks. Someone who complains about risk management when taking risk
27.
by Zolde 3y ago
You never had a point.
28.
by Zolde 3y ago
It feels like a little too late. Science has to wait on controls, and that is difficult to do with a new disease. Civilian science/shitposters were promoting tea and catechins since early 2020, but such information was classified as da
29.
by Zolde 3y ago
Thanks and interesting. You can use your reply to find similar inventions, but completely impossible for me to judge on sense. "I've been exploring the applications of p-adic analysis in quantum mechanics, especially as they perta