snewman

11 ms·

91.
Transformative AGI by 2043 is <1% likely(arxiv.org)
33 points by snewman 3y ago|41 comments
92.
Why We Won't Achieve AGI Until Memory Is a Core Architectural Component(amistrongeryet.substack.com)
3 points by snewman 3y ago|0 comments
93.
by snewman 3y ago
Yes, although a better way of saying it is that it is not the helium which has lifting power, it is the balloon. A balloon has lifting power equal to the weight of the air displaced by the volume of the balloon. From this, you must subtract
94.
Process vs. Product: Why We Are Not yet on the Cusp of AGI(amistrongeryet.substack.com)
3 points by snewman 3y ago|0 comments
95.
by snewman 3y ago
I maintain two, with pretty different content. Both are intended to explore deep topics (AI and climate change, respectively), aiming for a middle ground between academic papers (meaty, but often hard to understand + contextualize if you&#x
96.
by snewman 3y ago
Regarding opportunity cost, it depends on what your limiting factors are. For instance, to the extent that solar installations in the US are currently limited by grid interconnect, this is a good way to deploy additional solar in service of
97.
Georgia program is training a cleantech manufacturing workforce(canarymedia.com)
1 points by snewman 3y ago|0 comments
98.
by snewman 3y ago
Great post. I'll propose one more "myth": writing a blog doesn't benefit you personally, except by drawing attention . I've found that writing about a topic can also be a great way to solidify your own thinking on
99.
by snewman 3y ago
So much negativity on this thread. I think it's at least worth mentioning that the paper does present many genuinely interesting ideas; for instance: 1. The change history of a repository contains more information than just the final s
100.
To Address AI Risks, Draw Lessons from Climate Change(amistrongeryet.substack.com)
1 points by snewman 3y ago|3 comments
101.
by snewman 3y ago
This is oversimplified at best. Not all VCs are created equal. A broad-brush characterization like this removes any incentive for VCs to behave well (if everyone will assume the worst of you, why bother trying to build a good reputation), a
102.
A Unified Theory of AI Risk(amistrongeryet.substack.com)
2 points by snewman 3y ago|0 comments
103.
by snewman 3y ago
I think the value here is around starting with a known default . If someone hands you a contract which consists of a known boilerplate plus three changes, then (assuming you're already familiar with the boilerplate) you don't nee
104.
by snewman 3y ago
You are correct that water vapor is the primary greenhouse gas. Despite that, water vapor emissions are not generally considered to impact warming. I think that is because the water content of the atmosphere is already in balance due to the
105.
Get Ready for AI to Outdo Us at Everything(amistrongeryet.substack.com)
2 points by snewman 3y ago|0 comments
106.
An Intuitive Explanation of Large Language Models(amistrongeryet.substack.com)
2 points by snewman 3y ago|0 comments
107.
I’m a Senior Software Engineer. What Would GPT-4 Need to Do My Job?(amistrongeryet.substack.com)
3 points by snewman 3y ago|0 comments
108.
by snewman 4y ago
[author here] It's true that I didn't present a lot of concrete evidence of GPT-4's limitations. This is a blog post, not an academic paper. To my mind, the most concerning thing about GPT-4's performance on these two ex
109.
by snewman 4y ago
I don't think this sort of dismissal adds anything to the conversation. I myself, in the act of typing this comment, am "generating text". I think it's interesting to discuss the capabilities of leading-edge language mod
110.
How 8 billion people warmed 5,500,000 gigatons of air(climateer.substack.com)
1 points by snewman 4y ago|0 comments
111.
Visualizing CO₂(climateer.substack.com)
6 points by snewman 4y ago|0 comments
112.
by snewman 4y ago
These bonds were supposed to be "safe" in the sense that they would be repaid on schedule. And that's been fine, AFAIK there is no suggestion that they won't be repaid. The problem is that SVB, knowingly, took on "i
113.
by snewman 4y ago
1.5% of $80B is actually $1.2B
114.
by snewman 4y ago
There are many, many uses for green hydrogen unrelated to road vehicles. Long-term power storage (use cheap summertime solar to bank hydrogen for utility power in the winter), ammonia synthesis for fertilizer, replacing fossil fuels for hig
115.
by snewman 4y ago
I don't think we ever developed any software targeted at the Lisa (or if we did, it was in the brief period before I joined). We were always developing for the Mac, initially cross-compiling from Lisa. We did move to Mac-based developm
116.
by snewman 4y ago
I can confirm this regarding Mac development. I was part of the team that built FullPaint ( https://en.wikipedia.org/wiki/FullPaint ) back in 1985. We had just a few Lisas to work with, and they were considered sacred. T
117.
by snewman 4y ago
This (STEPS) sounds interesting, do you have a pointer? (It's not the easiest word to google...)
118.
by snewman 4y ago
[Author here] The LUIE figure we found for dedicated biomass is 580 m²/MWh/y: 580 square meters to generate one megawatt-hour per year. You multiplied this by hectares, meaning you multiplied area times area. If you want to determ
119.
by snewman 4y ago
> Maybe we could grow corn for food, but cram the stalks into the generator and sequester the remaining carbon? This is a thing; it's called "BECCS". You can also sequester the carbon without trying to burn it first; see,
120.
by snewman 4y ago
> woody biomass isn't coming from clear cutting forests, its from necessary thinning to keep forests healthy From what I've seen, claims to this effect often turn out to be false. For instance, from https://billmckib

More ›