Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
dqpb
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
19 ms
·
151.
▲
by
dqpb
4y ago
They mostly don't say nonsense. I listened to the whole thing and I found it very convincing. I've never been able to listen all the way through AI generated content before. It's hard for me to believe this was 100% AI genera
152.
▲
by
dqpb
4y ago
This is the problem with "fake news" and increasingly AI generated content. From an information theoretic perspective this is a perfect title. It sounds 100% improbable, it is therefore 100% surprising, yet it (apparently) deliver
153.
▲
by
dqpb
4y ago
This is amazing.
154.
▲
by
dqpb
4y ago
People are acting like 20% of Twitter content is crap and it can be saved by moderating that 20%. This is incredibly naive. 99% of Twitter is crap and always will be. You can't moderate that away. The question is how to make the 1% dis
155.
▲
by
dqpb
4y ago
> by producing a lot of false, semi-false and nonsense content to drown out the signal in the noise. This is the natural state anyway because most people are idiots about most things.
156.
▲
by
dqpb
4y ago
My main problem with Go is the error handling.
157.
▲
by
dqpb
4y ago
I'm using the word as broadly as it's meant to be used.
158.
▲
by
dqpb
4y ago
You should look up the word in a dictionary. You might be surprised by its definition.
159.
▲
by
dqpb
4y ago
So, if I already model systems via boxes and arrows, and use things like petri nets, then am I good, or do I still need to learn category theory?
160.
▲
by
dqpb
4y ago
Those are technical problems. "Technical" doesn't just mean tech stack.
161.
▲
by
dqpb
4y ago
Massive Multiplayer Online... something?
162.
▲
by
dqpb
4y ago
You agree and then list a bunch of technical problems.
163.
▲
by
dqpb
4y ago
I think your looking at the problem wrong. It's not Twitter's job to fix people. It's Twitters job to connect people in a communications graph. This is absolutely a technical problem, as there are many possible configurations
164.
▲
Neural Correlates of Consciousness
(en.wikipedia.org)
4 points
by
dqpb
4y ago
|
0 comments
165.
▲
by
dqpb
4y ago
As someone who has always thought Twitter is stupid, I'm glad Musk bought it and is shaking things up. Both possible outcomes are interesting - either Twitter finally becomes not stupid, or even the worlds richest man and greatest tech
166.
▲
by
dqpb
4y ago
Can't you solve the moderation problem by simply having the default view be the social graph of depth N of people you subscribe to? In that case, unwanted comments actually provide useful information about your social graph.
167.
▲
by
dqpb
4y ago
Sex is the "meaning of life".
168.
▲
by
dqpb
4y ago
The only gift I request is a truly novel idea.
169.
▲
by
dqpb
4y ago
> the IBM salesmen stood around talking to each other. Whether or not this actually sold computers was a question we never discussed.
170.
▲
by
dqpb
4y ago
Docker is an interesting thing. I really love what I'm able to do with it. But I despise when people make it a necessary component of the development environment.
171.
▲
by
dqpb
4y ago
Sounds like a win-win!
172.
▲
by
dqpb
4y ago
I'm totally in favor of bringing back dinosaurs. I'm also in favor of modifying animals to have more human-like brains. I understand that many/most people would oppose this on ethical grounds. I could probably even articulate
173.
▲
by
dqpb
4y ago
Money is decision-making-tokens. I believe the best people to use those tokens are the people who earned them. But, when a huge portion of societies decision tokens are sitting idle in a few accounts, it calls into question whether the hold
174.
▲
by
dqpb
4y ago
> Modern machine learning systems are not based on biological ... rules ... where you could uphold a principle of noncontradiction
175.
▲
by
dqpb
4y ago
> Your response is the playbook for these types of scams. > That doesn’t mean that the long running Google work in the field, etc is a scam. That doesn’t mean that working on self driving is unethical. Seems like you're still gra
176.
▲
by
dqpb
4y ago
The technology has come a long way since the early days of the DARPA grand challenge. Fortunately, the technology will continue to progress, independent of your ability to comprehend it.
177.
▲
by
dqpb
4y ago
It's not a scam. A scam is a deliberate deception. Engineers working on this are genuinely trying to solve the problem.
178.
▲
by
dqpb
4y ago
Or lead
179.
▲
by
dqpb
4y ago
It would be nice to see a smoker-adjusted rate.
180.
▲
by
dqpb
4y ago
> need some society-scale effort to nail down I think a good approach would be for people to build things that exhibit consciousness according to whatever their model is and claim "this is conscious". Then let people debate whe
More ›