jedharris

9 ms·

1.
State-of-the-Art Performance on AMD MI355 – In 14 Days(modular.com)
6 points by jedharris 11mo ago|0 comments
2.
The only AI curve that matters(exponentialview.co)
2 points by jedharris 1y ago|2 comments
3.
by jedharris 1y ago
Never mind debates about intelligence. How many actions can a system take at 99% reliability before a human must intervene? This metric is doubling every 7 months.
4.
Pretraining Under Infinite Compute(arxiv.org)
3 points by jedharris 1y ago|1 comments
5.
by jedharris 1y ago
"Our results show that simple algorithmic improvements can enable significantly more data-efficient pre-training in a compute-rich future."
6.
by jedharris 1y ago
Certainly one of my criteria for a utopia would be that some of the beings in it are engaged in intriguing or exciting activities with meaningful consequences. I'd require more than Special Circumstances to meet this criterion -- they
7.
by jedharris 1y ago
In the Culture Series history hasn't ended, and the members of the Culture can and do leave and live in ways the Culture would find repugnant. Also there are marginal participants (such as the Mind known as Meat Fucker) who are more or
8.
by jedharris 1y ago
This is an example that supports Scott's point that people don't have world models. The people who "believe" this don't wonder how stock market continues to operate now that NYC is a wreck. Etc. I wonder in what sen
9.
by jedharris 1y ago
> it's possible for simple normal trains of thought to latch your brain into a very undesirable state. This seems very incorrect, or at least drastically underspecified. These trains of thought are "normal" (i.e. common an
10.
by jedharris 1y ago
some pointers to the research program please?
11.
Tolerance is a social contract not a moral precept(medium.com)
3 points by jedharris 2y ago|1 comments
12.
by jedharris 2y ago
We often weaken ourselves by thinking we have a moral duty to be tolerant, even in the face of intolerance.
13.
by jedharris 2y ago
What the journey from AI assistant to full-virtuality can teach us about the nature of love.
14.
Gentle Romance(asimov.press)
3 points by jedharris 2y ago|1 comments
15.
by jedharris 2y ago
See also independent RL based reasoning results, fully open source: https://hkust-nlp.notion.site/simplerl-reason Very small training set! "we replicate the DeepSeek-R1-Zero and DeepSeek-R1 training on small models wit
16.
by jedharris 2y ago
They replicate the DeepSeek-R1-Zero and DeepSeek-R1 training on small models with limited data. We show that long Chain-of-Thought (CoT) and self-reflection can emerge on a 7B model with only 8K MATH examples, and they achieve surprisingly
17.
by jedharris 2y ago
PDF available at < https://dare.uva.nl/search?identifier=41ed9d6d-70a7-4bc7-882... >
18.
by jedharris 2y ago
None of the MP4 videos play in Firefox. The MOV videos play fine. All the videos play in Safari. I have not tried Chrome.
19.
Solving the long tail with end to end AI(wayve.ai)
1 points by jedharris 2y ago|0 comments
20.
by jedharris 3y ago
That matches how philosophers typically talk about the Chinese room. However the Chinese room is supposed to "behaves as if it understands Chinese" and can engage in a conversation (let us assume via text). To do this the room mus
21.
by jedharris 3y ago
Very fully open-sourced (hardware, software, data). About $30 K BOM. Seems like a great starting point for open source human scale robotics.
22.
by jedharris 3y ago
Also, you are obviously wrong (or rhetorical?) when you say "No one is talking about what a machine intends to do." We certainly do! You can say "No one should" or other normative condemnations but then we're arguin
23.
by jedharris 3y ago
Some philosophers take that position. Dennett, explicitly cited in the article, wrote The Intentional Stance (1987) about exactly the approach to intentionality taken in the article. His approach is accepted by many philosophers. As you p
24.
by jedharris 3y ago
Cool way of putting it. Let's run with that. A good actor can be seen as instantiating a Boltzmann Brain while on stage -- especially when improvising (as always may be needed). Maybe each of us is instantiating some superposition of B
25.
by jedharris 3y ago
The article provides a very clear reason for using the idea of "intention": that framing helps us understand and predict the behavior. In contrast framing a mathematical function as having "intention" doesn't help.
26.
by jedharris 3y ago
I'm surprised / disappointed this got so little attention on HN.
27.
by jedharris 3y ago
Very clean design, easier to understand than Rust (partly because better explained). Maybe shows the outlines of a clearer way to think about parallelism, which I hope will be fleshed out by further design moves, some of which are previewed
28.
by jedharris 3y ago
The reasons in the story are mostly symptoms of financialization -- the people making the major decisions were only looking at financial metrics, not operational metrics. The operational failures were driven by financial constraints. In tur
29.
by jedharris 3y ago
I'm surprised that none of these comments address the completely new interaction paradigm -- eye tracking + hand gesture tracking. Previous devices couldn't track accurately enough to provide high quality user interaction this way
30.
by jedharris 3y ago
The top comments miss the big driver for the first couple of years: business applications that are remote and/or collaboration heavy. I think these can easily drive more demand than Apple can fulfill in the first few years. Such applic

More ›