petra

9 ms·

151.
by petra 1y ago
But llms are relatively rarely used, and on the other hand, perf/latency is important to ux, and perf is variable(simple question, complex question, visual work). Those demand are better fullfiled at the cloud.
152.
by petra 1y ago
What determines whether 300k is scalable ?
153.
by petra 1y ago
This sounds brilliant. What's missing so you could create a demo for vc's or the relevant companies , proving the potential of this as competitive server-class core ?
154.
by petra 1y ago
There's a lof of innovation in optical interconnect, including at the chip-to-chip and die-to-die levels: https://www.eetimes.com/silicon-photonics-and-co-packaged-op...
155.
by petra 2y ago
For discovery of niche products,The Google search ads(without spying) system is a great solution. The issues of monopoly should be handled of course. And regarding word of mouth: Is word of mouth for great products really random?
156.
by petra 2y ago
Acquisition is a common result for sucsessful deep tech startups. It's because scaling in these markets is hard and better done at large companies.
157.
What if we could reason about the design space of data structures?(daslab.seas.harvard.edu)
1 points by petra 2y ago|0 comments
158.
by petra 2y ago
Maybe this would be helpful: http://daslab.seas.harvard.edu/datacalculator/
159.
by petra 2y ago
So pharmaceutical research is largely an engineering problem, of running experiments and synthesizing molecules as fast, cheap and accurate as possible ?
160.
by petra 2y ago
Assuming a decent AI and robotics, is a lot of working age population still a good thing? Or just more mouths to feed?
161.
by petra 2y ago
That Google product isnt that good, it can't really replace research done by a person.
162.
by petra 2y ago
There's work done on storing hydrogen in special liquids. If that works, there still are a lot of issues, but maybe it could fit heavy vehicles.
163.
by petra 2y ago
But the IP is already "compiled", or if you want to isolate even further, separate just the mask making process. What could be learned from masks?
164.
by petra 2y ago
Maybe bio weapons could be tested on organ-on-a-chip/body-on-a-chip systems?
165.
by petra 2y ago
You also need to be able to test your solution, on how sucsessful it is. In some domains it is harder than math and code.
166.
by petra 2y ago
What about using simulation(or estimation using software) to understand the manufacturing process well enough to design for it, without being in the factory?
167.
by petra 2y ago
There are plenty of industries where product engineering is done at a different company or place than product manufacturing. For example, consumer electronics, industrial machines and robotics, telecom and medical devices.
168.
by petra 2y ago
It doesn't have to have a use. It's an addictive form of enterntainment. Some like Twitter better, others like facebook better. It's just different ambience.
169.
by petra 2y ago
Modern packaging - high density 2.5D/3D is defintely not a commodity. Final packaging is.
170.
by petra 2y ago
Given that robot-as-a-service removes the biggest barriers for companies not buying robots, why aren't we seeing a huge growth in "employed" robots?
171.
by petra 2y ago
It does use knowledge from creators. But using knowledge from others is a big part of modern society, and the legal ways of protecting knowledge from commercial reuse are actually pretty limited. Is the result of an llm an accurate copy o
172.
by petra 2y ago
You can look at fully diagnosed cases(via surgery for example) and their previous scans.
173.
by petra 2y ago
I wonder if that will be the effect of the rp2350, to push the whole market toward software defined peripherals?
174.
by petra 2y ago
But given inference time compute, to give a strong reply reasonably fast, you'll need a lot of compute, very rarely used. Economically this fits the cloud much better.
175.
by petra 2y ago
I'm not sure the product is the real story here. It's probably just a way to sell their design process. I have no way of judging how valuable it is, but achieveng 7x improvement in efficiency of this wind turbine, and the team thi
176.
by petra 2y ago
The kagi experience isn't great, limited number of results, for example. I can't say I get a feeling they are much better than Google day to day. But the fact you can use kagi to filter out links with ads/trackers inside them
177.
by petra 2y ago
There are already some VC funded startups in the field of piano teaching and since piano is a digital tool, they are probably working on including every part of teaching, besides the interpersonal connection between student and teacher. I w
178.
by petra 2y ago
Maybe it's rare to do curiosity driven research. But from the days of Bell Labs, haven't we greatly improved our ability to connect between some research concept to the idea of doing something useful, somewhere ? And once you hav
179.
by petra 2y ago
Yes, it was basic research( guided to the field of machine learning), but between a search monopoly and their autonomous car project, they definitely have a great economic engine to use that basic research and the talent it pulled into Goog
180.
by petra 2y ago
Sure, news agencies aren't a way to guarrantee truth. That has always been a problem. But social networks have increased that problem by an order of magnitude. And we're seeing the results - the rise of dictatorships all around th

More ›