voidspark

8 ms·

1.
by voidspark 1y ago
Concerned about a "brand in the title"? What do you expect? The author is David Rosenthal, who was employee #4 at Nvidia (Chief Scientist). He's not some random historian or interviewer. That's his life experience.
2.
by voidspark 1y ago
The paid version "SuperGrok" has a larger context window, but nothing beats Gemini for that. I tried your question with SuperGrok. Here's the result. https://grok.com/share/bGVnYWN5_d298dd12-9942-411c-900
3.
by voidspark 1y ago
Michael Jackson was the biggest in history. Global megastar before the internet. Taylor Swift has a lot of sales but in terms of global significance and cultural impact there is no comparison. Not in the same league. Taylor Swift has relati
4.
by voidspark 1y ago
Those are instruments played and sequenced by humans. Meaningless comparison.
5.
by voidspark 1y ago
Comparing music and algorithms is nonsense. AI music is disposable generic soulless trash, even if it is technically correct, in accordance with the rules and conventions of music theory. AI generates Muzak. Totally generic and derivative.
6.
by voidspark 1y ago
This is exactly the problem. The top level executives are setting up to retire with billions in the bank, while the workers develop their own replacements before they retire with millions in the bank. Senior developers will be mostly obsole
7.
by voidspark 1y ago
It's his daughter. He is worried about his daughter first and foremost. Weird reply.
8.
by voidspark 1y ago
Yeah just being respectful to those 14 astronauts who died. They are worth mentioning. Nasa had major setbacks - not an "oopsie". Didn't mean to hijack the thread. Well done Voyager team.
9.
by voidspark 1y ago
That's not my point. The learned painful lessons and their success rate is high.
10.
by voidspark 1y ago
Yeah it's not "bonkers" or "insane". They learned the hard way. Painful lessons.
11.
by voidspark 1y ago
Two space shuttles exploded, killing everyone on board. https://en.wikipedia.org/wiki/Space_Shuttle_Challenger_disas... https://en.wikipedia.org/wiki/Space_Shuttle_Columbia_disaste...
12.
by voidspark 1y ago
I meant a verbatim exact copy of all documentation they have ever been trained on - which they are not. Neural networks are not designed for that. That's not how they encode information.
13.
by voidspark 1y ago
The default chat interface is the wrong tool for the job. The LLM needs context. https://github.com/marv1nnnnn/llm-min.txt The LLM is a problem solver but not a repository of documentation. Neural networks are not desi
14.
by voidspark 1y ago
It is impressive when it finds subtle errors in complex reasoning. But even the dumbest model will call you out if you ask it something like: "Hey I'm going to fill up my petrol car with diesel to make it faster. What brand of die
15.
by voidspark 1y ago
Theft or loss has always been a problem since life evolved on Earth. I don't think anyone claimed that crypto was un-losable or un-stealable. It's not magic. https://cryptosteel.com
16.
by voidspark 1y ago
Yes that's true, but no need to hold your crypto there as a permanent storage. Once your fiat is exchanged to crypto, immediately transfer the crypto to your private wallet.
17.
by voidspark 1y ago
I don't really care about crypto personally (volatile shitcoins) but I think that's a straw man argument. They all know it gets troublesome when it comes to dealing with fiat transactions. The hardcore crypto enthusiasts want to a
18.
by voidspark 1y ago
I have never been thanked for replying here before. Thanks. The optional choices happen when it tries to reason out a solution, but then finds it is making too many assumptions of unknown details about the user's system, preferences, g
19.
by voidspark 1y ago
This is an exchange problem, not a crypto problem. You don’t need an exchange to hold crypto.
20.
by voidspark 1y ago
IBM Watson is old, before deep learning. Deep learning systems have exceeded the hype. In 2016 we saw potential with models like AlphaGo Zero but no one could foresee the capability of LLMs (a type of deep learning model).
21.
by voidspark 1y ago
We don’t know if they had access to everything. They got data for “less than 1% of monthly transacting customers”.
22.
by voidspark 1y ago
No. Coinbase deals with fiat money, therefore subject to AML and KYC regulations.
23.
by voidspark 1y ago
Not if you are dealing with a regulated exchange that facilitates fiat money transactions. You can receive crypto privately to your own wallet without sharing PII, without any exchange.
24.
by voidspark 1y ago
How can customer support operate without knowing anything about the customer?
25.
by voidspark 1y ago
> "Do you want either [A, B, or C]" That's an example of what I'm talking about. Watch the reasoning process produce multiple options. That's what it is trained to do. That is problem solving, not "engagemen
26.
by voidspark 1y ago
Are you employed at Google or OpenAI? Are you working on these frontier models? In the case of medical questions it needs to know further details to provide a relevant diagnosis. That is how it was trained. In other cases you can observe it
27.
by voidspark 1y ago
I have been using it up to 100k tokens so far without issues. Never needed to go further than that. But much of that was in uploaded documents.
28.
by voidspark 1y ago
It's for higher quality of output. Better solutions. These are the state of the art reasoning models (subscription only, no free access) which are smarter. It also mainly happens when the context is clear that we are collaborating on w
29.
by voidspark 1y ago
> inability to self-reflect and recognize they have to ask for more details because their priors are too low. Gemini 2.5 Pro and ChatGPT-o3 have often asked me to provide additional details before doing a requested task. Gemini sometimes
30.
by voidspark 1y ago
http://aistudio.google.com

More ›