526 karma · joined December 30, 2010
Website: https://helbl.ing/
I've always been stronger at discrete type math/programming, which is why I tend to shy away from statistics-based stuff like ML.
One thing to note is that LLMs are indeed feed forward, however the generation of the text (from my understanding) is recursive in that you feed each output token to another forward pass of the neural network.
See the about section here: https://www.meetup.com/bostongamedev/
It's a real shame that BUG is dying, it was one of my favorite meetup groups prior to COVID.
Purely functional on the other hand has none of these problems. This is the approach taken by the Unison Language people, which I think makes the right design decisions.
Proof sketch: Turing Machines are enumerable by lexiographically sorting them. The Turing Machines that compute computable numbers are a subset of them. Therefore the set of computable numbers is subcountable.
Now assume that the set of computable numbers is enumerable by some Turing Machine M. Then I can construct a new computable number who's ith digit is a swap from the ith number's ith digit given by M (ie a diagonialization). This contradiction shows that the set of computable numbers is not enumerable.
Another way of looking at this is that determining if some Turing Machine computes a computable number requires deciding the halting problem, which cannot be done. TMs computing computable numbers are required to terminate given an input precision.
If machine learning makes inroads into pure mathematics, it will have to be from some sort of reinforcement learning, not supervised learning.
How often does Google release these new domains? I know that .ing was delegated to them in 2014... why are they allowed to control these domains if they never release them?
Remote: Yes
Willing to relocate: Yes, to within the US
Technologies: Programming languages, compilers, functional programming, formal methods, proof assistants, virtual reality, game development, simulation development, PyTorch, machine learning, research & development, computer graphics, computer networking, embedded systems
Resume/CV: https://github.com/calebh/calebh.github.io/blob/master/resum...
Email: caleb.helbling AT yahoo.com
--------------------
Computer science student living in Lafayette, IN. I am currently in the PhD program at Purdue University but am planning on leaving early with a master’s degree. Looking for a full time position after I graduate in May 2023.
I was most recently employed as a virtual reality software developer at Holos, Inc. My primary interests are in programming languages, computer graphics and machine learning.
In the past I was at MIT Lincoln Laboratory in the position of Assistant Staff. While there I primarily focused on embedded systems development, computer networking and graph theory research. I left this position in early 2019. Prior to this I was an undergraduate student, and held multiple internships at various companies. I have three publications in the areas of programming languages and graph theory.
So many Google products have been regressing lately. I've seen complaints about search, and GMail spam filters.
Meta programming is an excellent target, but it usually needs to be directly baked into the language. The closest system I've seen to a nice language agnostic close to meta programming system is Jetbrains MPS and Spoofax Language Workbench. Last time I checked, MPS is mostly dead and Spoofax is only used my programming language researchers.
For the Arduino programming stuff, I currently maintain a domain specific functional reactive programming language for Arduinos called Juniper: https://www.juniper-lang.org/
However the intersection between people who do Arduino programming and functional programmers is quite small. Also my interest has waned on and off over the years which means there have been long gaps between releases. I just work on it for fun mostly.
At Holos, we're developing a general-purpose spatial computing system for AR and VR, and deploying it in the educational and training fields to let people take control over their learning and personal growth like never before. Our product is built on top of proven technologies like Leap Motion's hand tracking, Valve’s SteamVR platform, Node.js, and Vue.
Our team is small but mighty and growing. You'll be working alongside the CTO, and often the Co-Founders, in a very generalist and multidisciplinary role. If you have an affinity for technology, design, and education, then please reach out. We welcome friendly, curious individuals that understand the power of communication, knowledge, and teamwork to apply.
We are currently seeking:
- AR/VR Interaction Engineer: https://angel.co/company/holosxr/jobs/1370420-unity-networki...
- Unity Networking Engineer: https://angel.co/company/holosxr/jobs/1370420-unity-networki...
At Holos, we're developing a general-purpose spatial computing system for AR and VR, and deploying it in the educational and training fields to let people take control over their learning and personal growth like never before. Our product is built on top of proven technologies like Leap Motion's hand tracking, Valve’s SteamVR platform, Node.js, and Vue.
Our team is small but mighty and growing. You'll be working alongside the CTO, and often the Co-Founders, in a very generalist and multidisciplinary role. If you have an affinity for technology, design, and education, then please reach out. We welcome friendly, curious individuals that understand the power of communication, knowledge, and teamwork to apply.
We are currently seeking:
- AR/VR Interaction Engineer: https://angel.co/company/holosxr/jobs/1370420-unity-networki...
- Unity Networking Engineer: https://angel.co/company/holosxr/jobs/1370420-unity-networki...
Amazon doesn't seem to be at all interested in tackling these issues - it's only a matter of time before it degrades completely into something like AliExpress.
I think that the best solution in Haskell would be to group the types in a typeclass. However then the typeclass is not directly associated with the p or img value constructors. I wonder if there is some way to use dependent types to improve this situation.