@kellogh@hachyderm.io
@kellogh@hachyderm.io avatar

kellogh

@kellogh@hachyderm.io

I'm a software engineer and sometimes manager. Currently #Raleigh but also #Seattle. Building ML platform for a healthcare startup. Previously, built an IoT platform for one of "those" companies.

Open source: dura, fossil, Jump-Location, Moq.AutoMock, others

Do I have other interests? No, but I do have kids and they have interests. I think that counts for something. I can braid hair and hunt unicorns!

I put the #rust in frustrate

He/Him

#metal #science #python

This profile is from a federated server and may be incomplete. Browse more on the original instance.

kellogh, to random
@kellogh@hachyderm.io avatar

rain is cathartic

kellogh, to random
@kellogh@hachyderm.io avatar

what’s the smallest group of people you’ve witnessed a reply-all email storm?

kellogh, to random
@kellogh@hachyderm.io avatar

a hard thing to fully embrace — stackoverflow was a phenomenal success for VCs, every way you look at it. it’s an example of the system working as intended

vancha, to llm Western Frisian
@vancha@fosstodon.org avatar

Looks like today I finally found a good application for 's: Learning languages!
I've been attempting to learn through duolingo for a while now, without much success. I figured if there's one thing language models should be good at, it's languages. So far the thing has actually been pretty helpful.

kellogh,
@kellogh@hachyderm.io avatar

@vancha i’ll be interested in how gpt-4o works. the main feature i’m interested in is “please slow down, i’m a n00b”

hgrsd, to random
@hgrsd@hachyderm.io avatar

Ok, fine, I heard you Rustaceans. I'll give Helix another shot.

kellogh,
@kellogh@hachyderm.io avatar

@hgrsd i love the idea of helix, but i can’t dedicate enough time to rewire my brain enough to actually love it

ianbicking, to random
@ianbicking@hachyderm.io avatar

I was looking over some of the new EU AI restrictions, and notably it barred educational institutions and employers from using biometric emotion detection.

This seems fine, but #gpt4o makes this complicated. You can't turn off its emotion detection, it's not even really explicit, but it's very much there. The demo where the person breathes hard and soft is one example, but GPT-V (vision) already showed how it could detect emotion in images and act on that.

kellogh,
@kellogh@hachyderm.io avatar

@ianbicking it seems like the obvious way to turn it off is to not use audio or vision, right? not a great solution, but it’s only for employers (presumably against their employees) and educators (presumably against their students)

kellogh, to random
@kellogh@hachyderm.io avatar

the news is cool, but now i want to see an embedding model that i can use with a streaming interruptible conversation

kellogh,
@kellogh@hachyderm.io avatar

also, now with #gpt4o, latency is going to be critical if you’re doing streaming audio/video, so #python may start looking less appealing. what’s the new #LLM language? #rust? #go? #cpp? #fortran?

#rustlang #golang

kellogh,
@kellogh@hachyderm.io avatar

there’s a lot of unanswered questions. certainly exciting times ahead (and a new batch of epic fails)

kellogh,
@kellogh@hachyderm.io avatar

@mistersql are you starting that debate? 😂

seriously though, idk, python is not my go to low latency language. so much of the language is written without performance in mind, e.g. a for loop raises a StopIteration error to stop looping. integers use 28 bytes, etc.

kellogh,
@kellogh@hachyderm.io avatar

@dr2chase your totally right. python has an enormous ecosystem, which makes it fantastic for slinging pipelines together. however, the announcement today says that the audio responds within ~350 milliseconds. i’d say your app can add 50-100 more without changing the experience much, but python apps typically add more like 200-400 ms

imo rust might be the next contender. they’ve been building out a lot of new data science & engineering tools with it. idk

kellogh,
@kellogh@hachyderm.io avatar

@dr2chase not that you couldn’t use python, it’s just that it doesn’t have great tools for being latency conscious

kellogh,
@kellogh@hachyderm.io avatar

i predict that there’s always going to be strong advantages to using for , but with streaming audio & video of , there’s not enough latency slack for python.

i think a framework will emerge, similar to pyspark, where you can write python code that gets compiled into a steaming plan, and executed as highly optimized low level code with the possibility of python UDFs. i figure it’s still a couple of years from being really usable rn

kellogh,
@kellogh@hachyderm.io avatar

i suppose keras & pytorch effectively do this for neural nets

kellogh, to random
@kellogh@hachyderm.io avatar

alright, is pretty cool, mainly bc it’s so gosh darn fast

kellogh, to random
@kellogh@hachyderm.io avatar

i watched a demo video of singing a song and, sure it was bad, but it was still better than Laura Trump

simon, to random
@simon@simonwillison.net avatar

My notes on this morning's OpenAI release of the new GPT-4o model - not a huge leap in "intelligence" (whatever that might mean) but still very significant thanks mainly to the impressive new audio capabilities and the drop in price - 50% cheaper via API, and should soon be available for free ChatGPT users as well https://simonwillison.net/2024/May/13/gpt-4o/

kellogh,
@kellogh@hachyderm.io avatar

@simon how do you think embeddings will fit into streaming audio & video? i can’t imagine RAG can go away, but certainly we need new embedding models, right?

kellogh, to random
@kellogh@hachyderm.io avatar

just had a conversation with my neighbor — he’s wearing an NRA tshirt and explaining that the pistol he has strapped to his hip is for shooting copperheads. And also he’s concerned that we leave our garage open too much and that copperheads might get in

kellogh,
@kellogh@hachyderm.io avatar

i’m probably just some hippy to him, but it doesn’t seem like guns are a great solution for snakes. copperheads are docile creatures, just use a turkey stick (everyone has one, right??) to nudge it off into the trees

kellogh,
@kellogh@hachyderm.io avatar

@jimfl well yeah, ofc that’s what i’m talking about. those SNAKES STOLE MY WIRING. i had to use my turkey stick to chase them away!

kellogh, to random
@kellogh@hachyderm.io avatar

i wish “type checking for infrastructure” was a thing

my code declares that there should be a S3 bucket that’s different from that other S3 bucket, etc. —> spin up the type checker, it reads APIs and verifies, “yep, this code should run fine”

kellogh,
@kellogh@hachyderm.io avatar

it seems like a lot of assumptions are lost between code & terraform & k8s, and there’s not a great framework for quickly validating

kellogh,
@kellogh@hachyderm.io avatar

@olafurw i had thought CDK was more of a terraform equivalent, am i wrong? e.g. for creating & synchronizing infra with a code base

i’m saying “type checking” to talk about some process you could run pre-deployment to check if the code is compatible with current infrastructure. doesn’t have to be types, but similar to types it encodes your assumptions and can be automatically verified. just felt like “typechecking” was a good analogy

kellogh,
@kellogh@hachyderm.io avatar

@brianknight @olafurw got a github link?

matthew_mullins_, to random

When pg_vector dropped for Postgres I said it would be a matter of time until vector search was just a standard feature of databases. It's available in OLAP platforms like Snowflake and BigQuery, and now it's come to workhorses like Cassandra too. Data Threads

kellogh,
@kellogh@hachyderm.io avatar

@matthew_mullins_ it’s still not as standard as i thought it would be

ErikJonker, to ai
@ErikJonker@mastodon.social avatar

Somehow the relevance of this research using LLM, AI to train robots is not fully appreciated i think. If we can train robots in simulated worlds and that learning can be applied in real world applications, it seems that learning for Robots has no problems with regard to enough trainingdata or am i missing something ? Also this could really accelerate the applications for all kind of tasks.
https://eureka-research.github.io/dr-eureka/
#AI #Robots #DrEureka #LLM

kellogh,
@kellogh@hachyderm.io avatar

@ErikJonker i don’t think it’s a full answer, but it certainly helps. it probably needs extra fine tuning help from humans

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • JUstTest
  • GTA5RPClips
  • DreamBathrooms
  • InstantRegret
  • magazineikmin
  • khanakhh
  • Youngstown
  • ngwrru68w68
  • slotface
  • everett
  • rosin
  • thenastyranch
  • kavyap
  • tacticalgear
  • megavids
  • cisconetworking
  • normalnudes
  • osvaldo12
  • ethstaker
  • mdbf
  • modclub
  • Durango
  • tester
  • provamag3
  • cubers
  • Leos
  • anitta
  • lostlight
  • All magazines
    •