gibsonf1

16 ms·

151.
by gibsonf1 4y ago
Does anyone have experience with https://www.brex.com/ ?
152.
by gibsonf1 4y ago
https://www.forbes.com/advisor/banking/checking/best-busines... TIAA bank is intereting with their CIDARS which breaks balances > 250K into CDs created in partner institutions.
153.
Ask HN: Best Startup Bank with low insolvency risk?
1 points by gibsonf1 4y ago|4 comments
154.
by gibsonf1 4y ago
Category theory seems to have no relation whatsoever to operational human concepts which may explain how unuseful category theory actually is.
155.
by gibsonf1 4y ago
Right, they just make all things up by design, no matter what you ask about. There is a statistical chance that a pattern of words output may correspond to fact but just as a good a chance not. The LLM's literally know nothing about
156.
by gibsonf1 4y ago
If you use the LLM to summarize the results, it makes things up by design. As soon as you introduce the LLM into search, you lose.
157.
by gibsonf1 4y ago
Actually, not very easy as the LLMs don't give you sources and literally make things up by design.
158.
by gibsonf1 4y ago
Perhaps a graph of this metric over time will help in understanding where things are headed. The moment choice is added, in that parents can choose where to send their kids, the schools will have no choice but to improve or fail.
159.
by gibsonf1 4y ago
Except for the subtle fact that autonomous driving doesn't work with current ml/dl technology, so really not an issue at all at the moment.
160.
by gibsonf1 4y ago
The LLM has no idea what words mean, so its quite impossible for them to reason about things. They just statistically mash up patterns of words from massive input with no clue whatsoever what those patterns of words actually mean in the re
161.
by gibsonf1 4y ago
You may be mistaking reasoning ability with sheer volume of more patterns to match on. You can only reason when you have awareness within which to reason, which these systems do not.
162.
by gibsonf1 4y ago
There aren't really any current advances outside of sheer scale of input in the models, and all the engineering and hardware around achieving that scale. And I think the point is no matter how much input data you give the ml/dl s
163.
by gibsonf1 4y ago
All ml/dl systems have no awareness - they just output based on input training - like a calculator outputs an answer. So what it means to model in awareness is what you are doing right now in reading this sentence. You take these word
164.
by gibsonf1 4y ago
The ability to model an object in awareness and its causality that corresponds to its space-time reality
165.
by gibsonf1 4y ago
Fully agree with this article. Our definition for intelligence: "Intelligence is conceptual awareness capable of real-time causal understanding and prediction about space-time."[1] [1] https://graphmetrix.com/trin
166.
by gibsonf1 4y ago
The key issue I think is the lie that this system is intelligent in any way, it simply isn't. And if people approach it that way, that it's just statistically fabricating word text patterns from massive input word text, it will re
167.
For Chat-Based AI, We Are All Once Again Tech Companies’ Guinea Pigs(wsj.com)
58 points by gibsonf1 4y ago|81 comments
168.
by gibsonf1 4y ago
It is truly remarkable that the distinguishing characteristic of human intelligence vs other animals, conceptual reasoning, is not mentioned. "concept" appears 0 times in this paper.
169.
Sensitive US military emails spill online(techcrunch.com)
2 points by gibsonf1 4y ago|0 comments
170.
by gibsonf1 4y ago
The fundamental flaw with this entire branch is the premise that consciousness is mathematical. Math is a tool of reason, not the cause. And once you adopt this premise, it makes all progress toward understanding reason impossible especia
171.
by gibsonf1 4y ago
Its a good point, but the system is just statistically outputting words from human use, so its really the darkness of the humans that used those words that were input that perhaps he is referring to.
172.
by gibsonf1 4y ago
Wow, it seems like a fully frightening idea to have an ml/dl system with no clue if its right or wrong about anything (as it knows literally nothing, including that fact) to have any kind of role in autonomous agency.
173.
From Zero to Hero with Solid (Fosdem23)(fosdem.org)
1 points by gibsonf1 4y ago|0 comments
174.
by gibsonf1 4y ago
Ironically, using ml/dl language models, which have only statistics and no intelligence, is a Google killer to the extent that Google uses AI and destroys their real useful search. The declining quality of Googel search seems to only
175.
by gibsonf1 4y ago
Maybe this is why Google search is so bad now "That’s why we re-oriented the company around AI six years ago —"
176.
“Jailbreak” Bypasses ChatGPT's Ethics Safeguards(futurism.com)
9 points by gibsonf1 4y ago|1 comments
177.
by gibsonf1 4y ago
Oh wow, absolutely false advertising "challenge incorrect assumptions", ChatGPT output can be absolutely 100% false or true or in between - ie fully unreliable and in no way remotely able to challenge anything as ChatGPT doesn&#x2
178.
by gibsonf1 4y ago
Wow, that is truly not a good classifier with success that low.
179.
by gibsonf1 4y ago
We have 1 collection with 4 doc types (attribute, state, event, process) and 14 relations between them, so a join concept makes no sense for us. I think the schema is the key issue in having success vs not. (Apache Solr - billions of docs)
180.
by gibsonf1 4y ago
We sure don't do it like that! :)

More ›