Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
tito
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
91.
▲
by
tito
2mo ago
lovely! I'm a fan of asynchronous LLM systems like this. There are times where I prefer slow, complete responses to immediate replies. For example there's a billboard on 101 for Poppy Bank offering 4% interest and I asked my daily
92.
▲
by
tito
2mo ago
Oh, is that how it works? So, when somebody says a model is running at X tokens per second, it means that the thinking process is running at that, and output tokens are much lower then? Thanks to the explanation.
93.
▲
by
tito
2mo ago
Are there any well thought through stories about what this would look like? For example, I'm thinking about like nutrient flow, decision making, energy input, gravitational force, things like that seem to govern the value and speed of
94.
▲
by
tito
2mo ago
Done, Mac -> Max
95.
▲
by
tito
2mo ago
I commented similarly below, but as a terrible programmer, I probably perform about 1 minute per token too (at Kimi 3 level). It puts into context how I think about intelligence
96.
▲
by
tito
2mo ago
Having the right type of interface makes a huge difference. It reminds me of when Willow Garage chose to name their bot the TurtleBot, because if they named it anything else, people would think it was fast and capable. But when they called
97.
▲
by
tito
2mo ago
1 button Kimi morse code interface
98.
▲
by
tito
2mo ago
The Github specifically mentions an option to stream it from an online host. It's extremely slow.
99.
▲
by
tito
2mo ago
I like seeing the latest and greatest model crammed into new systems to see how it fares. To deal with the speed, one person on reddit suggested using it in an email interface rather than a chat interface.
100.
▲
by
tito
2mo ago
wow cool! I like watching the new models come out and how they end up crammed in to run on local machines. I learned what mxfp4 is thanks to this latest Kimi release - although it sounds like it means that there's less room for compres
101.
▲
Running Kimi K3 on a M1 Max
(github.com)
138 points
by
tito
2mo ago
|
93 comments
102.
▲
by
tito
6mo ago
Hey, it's Saturday night and I'm on Hacker News woot woot. I work on climate technology (sucking carbon dioxide out of the sky), and I have a side quest to create a "Freedom to Breathe" mural in Manhattan before the upco
103.
▲
by
tito
6mo ago
This is great. I'm typing this message now using Ghost Pepper. What benefits have you seen from the OCR screen sharing step?
104.
▲
by
tito
10mo ago
Build economies around carbon removal, and investing in climate interventions like sunlight reflection to buy time.
105.
▲
by
tito
10mo ago
Iceland has the 5th highest GDP per capita in the world. We’re about to witness what rich countries do when confronted by a changing climate.
106.
▲
by
tito
11mo ago
A tree eats about twenty kilograms of carbon dioxide per year. But burning a tree releases about 100 times that amount of carbon dioxide. So if just 1% of a forest burns for any reason, that reverses the rest. And if it’s more than that, th
107.
▲
by
tito
11mo ago
I just got back from the Millennium Seed Bank in the UK and was marveling at the size of a small footprint facility that stores samples of more than 10% of all known living plants. Reading this thread, I was curious about what the size of C
108.
▲
by
tito
1y ago
Hey, from the demo video, I thought it was entering the answers to your pre-entered questions based on the transcript, but it looks like you're just typing in the answers. Is that right?
109.
▲
by
tito
1y ago
It literally says close the facility in the budget. It's a little shed on top of the mountain. Here's a picture: https://education.nationalgeographic.org/resource/mauna-loa-...
110.
▲
by
tito
1y ago
It's an atmospheric monitoring station and a solar telescope. I've been there. it's basically a shipping container full of sensors and computers and two people huddled there keeping it running. The big telescopes are on Mauna
111.
▲
by
tito
1y ago
Seconded. The facility on Mauna Loa is the Mauna Loa Observatory which is atmospheric sensors and a solar telescope. The famous telescopes are on Mauna Kea (source: I grew up in Hawaii and worked for one of the telescopes, CFHT)
112.
▲
by
tito
1y ago
Catching up here. I visited there as part of building The State of Carbon website ( https://carbon.datahub.io/ ). This was much upvoted project on HackerNews back when it was first created: https://news.ycombinator
113.
▲
by
tito
1y ago
The current best in class methodology for biochar aims for 100+ years. There are some folks saying there are pathways to make biochar last 1000+ years. Puro methodology: https://biochar.groups.io/g/main/attachment&
114.
▲
by
tito
1y ago
The potential benefit of this idea is it's very cheap. One penny per ton vs hundreds of dollars of per ton of biochar. Iron fertilization, shading the sun, and more nukey-stuff, all worth exploring at this point.
115.
▲
by
tito
1y ago
Climate science consensus is we need to decarbonize and remove. We've gone too far at this point, decarbonizatin alone is no longer sufficient. So we need all the solar-y wind-y nuclear-y stuff AND all this weird removal stuff. And som
116.
▲
by
tito
1y ago
There is no climate model for less than 2 degrees of warming with decarbonization alone. It's unfortunate but that's where we are today. Large scale decarbonization and removal together are both required, and fast.
117.
▲
by
tito
1y ago
fascinating. Ignoring the ecological questions, I admire the "outside the box" thinking. For the sake of the calculation: $10^10 for 10^12 tons is an implied cost of one cent per ton. So three orders of magnitude cheaperish than c
118.
▲
by
tito
2y ago
I looked up my favorite mud awhile back. Just this post of other people looking for it for over 20 years! https://groups.google.com/g/alt.mud/c/Hn5EOqDfcyg
119.
▲
by
tito
2y ago
Agreed, we have to find a way to do both. It's unlikely we will prevail. As is, the path we're on is about 4 million tons per year of removals by 2030. A new model from Rocky Mountain Institute suggests we need to be at more like
120.
▲
by
tito
2y ago
Perhaps in 2010 this would have been sufficient. Unfortunately decarbonization alone is no longer enough to stabilize at 2 degrees of warming. For a stable climate massive decarbonization and removal is needed.
More ›