Silverstream | https://www.silverstream.ai/ | Founding Software Engineer, Senior Software Engineer | Full-Time | Remotely(Initially) then On-Site | San Francisco Bay Area
Silverstream is a fast-growing early-stage startup democratizing AI agents. We’re developing a cloud platform to create production-ready, in-house, browser-based AI agents. We’re at an inflection point in our growth and moving quickly. We’re in search of a passionate software engineer who will help scale our platform and influence the core technology. This position can be tailored to either a Senior Software Engineer or a Founding Software Engineer, depending on your experience and career goals.
For more details and to apply, please check the links below:
>genetic algorithms combined with RL where the genetics determine the reward function.
I have been working on this problem for years (2+ as researcher, 2 as PhD student).
The main issue is that evolution is both massively parallel and had plenty of runtime to get to human level intelligence.
The person that pushes this evolution/evolved reward point is Andrew G. Barto and his students/collaborators over the years.
Satinder Singh in particular is actively working on gradient based algorithms to find rewards (e.g. https://arxiv.org/abs/2102.06741)
> Maybe we need to frame RL goals in much more simple terms, and allow genetic algorithms to evolve their own inputs and reward functions on their own.
I was checking HN while the current iteration of this (gradient based, genetic was my master thesis) algorithm, the main complexity is figuring out:
1) What are the sub-goal e.g. grasping things
2) How to solve those goals e.g. motor control
3) How to do something useful, e.g. surviving
Balancing those three processes is the current hurdle.
Also, evolution isn’t trying to get to human level intelligence. It’s just one out of millions of adaptations that work, it’s recent, and it’s rare. Change Earths parameters a little over the past several million years, and maybe we don’t evolve.
> The main issue is that evolution is both massively parallel and had plenty of runtime to get to human level intelligence.
How many entities are we talking about for substantial evolution? I know that there have been 100 billion "humans" (not that it's so clear-cut) alive, so guessing this is on the order of ~trillions of entities to simulate some evolution for (but maybe I'm really underestimating the early tail of tons and tons of microorganisms and small short-lived life that got us to this point).
Is the bandwidth of evolution that much larger than what we could possible simulate with computation, especially for a much simpler world/task than "generally survive"?
The growth of the plant follows Space Colonization Algorithm which have been used for rending realistic trees in games. This algorithm is based on a cloud of points which have been inspired by the grown of tree in order to provide a certain attraction to the growing branches.
physical buttons are hard to debug!
emergency-only cellphone looks like the better alternative for now even if it would require me to use two phones or swap sims.
It's not a life risk situation, more a (business and peace of mind) life risk situation so i don't think ~life alert could work
Take a look at particle.io. They have a super simple button that you can connect via WiFi or GSM to ifttt or twilio really easily. Sample code leaves almost nothing to debug.
Thanks for the suggestion, I didn't think of Twilio=>webook.
Telegram is one of the platforms I use daily and something that would distract me but i guess you can replace it with anything else like an IRC client.
Agents are digital manufacturing machines and benefit from the same processes we identified for reliability in the real world