Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
johnnyb_61820
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
15 ms
·
91.
▲
Build Your Own Chatbot
(mindmatters.ai)
5 points
by
johnnyb_61820
7y ago
|
0 comments
92.
▲
Even Elon Musk Can't Get a Robotaxi These Days
(mindmatters.ai)
4 points
by
johnnyb_61820
7y ago
|
0 comments
93.
▲
by
johnnyb_61820
7y ago
"I believe Elon is stating that the liability may land on the car owner’s insurance" There is never liability on insurance. Insurance is just a method of payment. It is possible that insurance will cover this, but that's ir
94.
▲
Elon Musk: You Are Liable for My Malfunctioning Code
(mindmatters.ai)
12 points
by
johnnyb_61820
7y ago
|
5 comments
95.
▲
Facebook “Likes” Cryptocurrency
(mindmatters.ai)
2 points
by
johnnyb_61820
7y ago
|
0 comments
96.
▲
Machine Learning Tip: Set Boundaries for the Problems
(mindmatters.ai)
2 points
by
johnnyb_61820
7y ago
|
0 comments
97.
▲
True Believer Loses Faith in Fully Self-Driving Cars
(mindmatters.ai)
20 points
by
johnnyb_61820
7y ago
|
3 comments
98.
▲
by
johnnyb_61820
7y ago
Solomonoff induction has long been known to be important in machine learning, where the shorter of two equivalent models is considered "better". This paper extends this in three ways. First, it shows that the data itself can be
99.
▲
Generalized Information: A Method for Judging Machine Learning Models
(journals.blythinstitute.org)
1 points
by
johnnyb_61820
7y ago
|
1 comments
100.
▲
Machine Learning Dates Back to at Least 300BC
(mindmatters.ai)
2 points
by
johnnyb_61820
7y ago
|
0 comments
101.
▲
by
johnnyb_61820
7y ago
I'm curious why it is assumed by so many that it is certain that a sufficiently intelligent machine could produce even better machines. This is certainly at least not demonstrated, and all evidence I've seen for it is extremely w
102.
▲
A Critic of the Evangelical Statement on AI Misunderstands the Issues
(mindmatters.ai)
2 points
by
johnnyb_61820
7y ago
|
0 comments
103.
▲
Need Cash Fast? Just Pretend You Wrote Software
(mindmatters.ai)
6 points
by
johnnyb_61820
7y ago
|
0 comments
104.
▲
by
johnnyb_61820
8y ago
The fact that there was a well-known kludge to get around the problems inherent in the notation doesn't make the notation right. As the paper pointed out, the new notation didn't drop magically from the sky, it is literally the a
105.
▲
by
johnnyb_61820
8y ago
Well, I guess that's debatable. However, the way I see it, the old version was presented as a fraction, but couldn't be used as a fraction. The new version is also presented as a fraction and can be used as a fraction. Perhaps
106.
▲
Walter Bradley Center Fellow Discovers Longstanding Flaw in Elementary Calculus
(mindmatters.ai)
12 points
by
johnnyb_61820
8y ago
|
6 comments
107.
▲
Self-Driving Cars Need Virtual Rails
(mindmatters.ai)
1 points
by
johnnyb_61820
8y ago
|
0 comments
108.
▲
I Doubt That AI Can Match the Human Mind
(mindmatters.ai)
2 points
by
johnnyb_61820
8y ago
|
1 comments
109.
▲
by
johnnyb_61820
8y ago
I found it interesting that self-driving made the "backup" driver sufficiently inattentive that it rendered their help ineffective. Reminds me a little of the book "The Efficiency Paradox".
110.
▲
AI Ethics and the Value of Human Life
(mindmatters.ai)
2 points
by
johnnyb_61820
8y ago
|
1 comments
111.
▲
by
johnnyb_61820
8y ago
For those wanting a TL;DR: The main suggestions by the author are in section 2. Essentially, it says that modularity allows any errors to be contained, conserved design-by-contract interfaces allows for system expansion, and encapsulation
112.
▲
The Logical Possibility of Halting Oracles
(journals.blythinstitute.org)
2 points
by
johnnyb_61820
8y ago
|
0 comments
113.
▲
Navigating the Machine Learning Landscape, Pt. 2: Supervised Classifiers
(mindmatters.ai)
2 points
by
johnnyb_61820
8y ago
|
0 comments
114.
▲
Identifying Artificial Intelligence Blind Spots
(news.mit.edu)
1 points
by
johnnyb_61820
8y ago
|
0 comments
115.
▲
Remember the Luddites
(mindmatters.ai)
3 points
by
johnnyb_61820
8y ago
|
0 comments
116.
▲
Navigating the Machine Learning Landscape, Pt. 1
(mindmatters.ai)
2 points
by
johnnyb_61820
8y ago
|
0 comments
117.
▲
Pursuing Nerd Quality Over Nerd Quantity
(mindmatters.ai)
1 points
by
johnnyb_61820
8y ago
|
0 comments
118.
▲
How Can Information Theory Help the Economy Grow?
(mindmatters.ai)
3 points
by
johnnyb_61820
8y ago
|
0 comments
119.
▲
Google's Strength Is Harnessing Intentionality, Not Big Data
(mindmatters.ai)
2 points
by
johnnyb_61820
8y ago
|
0 comments
120.
▲
Be Choosy About What You Automate
(mindmatters.ai)
15 points
by
johnnyb_61820
8y ago
|
1 comments
More ›