J.R. Arredondo

Tweets about ML, cloud, product management, business, math. At Google working on conversational AI/ML. Ex-AWS, Ex-Solarwinds, Ex-Rackspace, Ex-Microsoft.

[J.R. Arredondo @jrarredondo]

@caytlyn_brooke Reminded me of this story. I hope it helps: https://t.co/ywmkvqAWZt

[Biola University @biolau]

We’re so proud of our alumnus Nate Norell & crew, made up of Biola students & alumni who created this commercial! We love Doritos & the commercial! Congrats, Abduction film crew! This represents the extraordinary talent from our film school, Snyder School of Cinema & Media Arts.

[J.R. Arredondo @jrarredondo]

@dvassallo Tomorrow:

[J.R. Arredondo @jrarredondo]

@AbdullahMdKhan @svpino What country are you in?

[J.R. Arredondo @jrarredondo]

@YitziLitt @docmilanfar @maciejhalber https://t.co/mQxGxxew8q

[J.R. Arredondo @jrarredondo]

TIL that ML is related to thermodynamics concepts such as entropy, free energy, and, wait for it, temperature.

[comonoidal esotericist 🇨🇿🇪🇺🇺🇸🦀 @adamnemecek1]

The entire field of machine learning exists because the partition function of the Boltzmann distribution is computationally intractable.

[Vaibhav (VB) Srivastav @reach_vb]

MARS5 TTS: Open Source Text to Speech with insane prosodic control! 🔥

> Voice cloning with less than 5 seconds of audio
> Two stage Auto-Regressive (750M) + Non-Auto Regressive (450M) model architecture
> Used BPE tokenizer to enable control over punctuations, pauses, stops etc.
> AR model predicts L0 coarse tokens, refined further by the NAR DDPM model followed by the vocoder

Great job Camb AI team! Kudos for open sourcing the artifacts - looking forward to what comes next ;)

[Alex Bodner @AlexBodner_]

🔥Convolutional Kolmogorov-Arnold Networks Preprint out now🔥
I'm very excited to release our preprint on Arxiv. Also, my first paper. We proposed a novel architecture that shows potential to compete with CNNs.
Want to learn more about it?
🧵Follow this thread👇

[J.R. Arredondo @jrarredondo]

"Any sufficiently advanced agent is indistinguishable from a DSL" (@Madrox, I think)

[caesararum, BS, DOGS @caesararum]

"i have 10 years experience building stable, reliable, maintainable software systems"

the software systems:

[J.R. Arredondo @jrarredondo]

@paulg This reminded me of this chart I saw in business school of earnings management by US companies (from “Earnings Management to Avoid Earnings Decreases and Losses” by Burgstahler and Dichev)

[Tim Rocktäschel @_rockt]

Like everyone else I am extremely excited and optimistic about AI progress. But since there seems to be the misconception that it is really easy to make predictions about the future for exponential trends (let's just draw lines on a log scale), this is a reminder that while observing an exponential trend (e.g. compute or neural network parameter increases), we can never be fully certain that what we are observing isn't in fact following a logistic trend before the inflection point. Before the inflection point, both trends look the same.

[ₕₐₘₚₜₒₙ @hamptonism]

Convolutional Neural Networks in action

[J.R. Arredondo @jrarredondo]

@dvassallo We have your actual name, we just let it all get leaked, but we still say hi with a dry “Hello.” At least say “Hello Daniel” since it’s already leaked anyway.

[J.R. Arredondo @jrarredondo]

We know ML sometimes works, but we often just don’t know why it works!

[Tanishq Mathew Abraham, Ph.D. @iScienceLuvr]

A reminder that no one knows what they're doing in deep learning and why things work and that the best we can do is pray for our models to perform 🤣

[Joscha Bach @Plinz]

What makes the whole LLM/self-awareness debate so tricky is that neurons are not self-aware either. Instead, our neurons *simulate* a person, which experiences itself as self-aware. Can LLMs simulate a self-aware person too, or are they just simulating the simulation of a person?

[J.R. Arredondo @jrarredondo]

This is fascinating! “Hyperparameter landscapes are fractal”

[Jascha Sohl-Dickstein @jaschasd]

Have you ever done a dense grid search over neural network hyperparameters? Like a *really dense* grid search? It looks like this (!!). Blueish colors correspond to hyperparameters for which training converges, redish colors to hyperparameters for which training diverges.

[J.R. Arredondo @jrarredondo]

… and now I cannot stop thinking about that. “Why neural networks and brains learn similar features” from @twimlai https://t.co/kRa9VYAv3i

[J.R. Arredondo @jrarredondo]

What’s the symbolic representation of intelligence? @naturecomputes: “all relevant information is encoded. Our brain transforms photons that hit retina or oscillations of air pressure that hit cochlea into electrical activity that somehow has rich perceptual cognitive experience”

[The TWIML AI Podcast @twimlai]

Efficiency is a fundamental principle in neuroscience. 🧠 Biological systems are resource-constrained and it is vital to transmit only what is essential!

Head over to https://t.co/BrgECo6TME to watch Sophia Sanborn’s (@naturecomputes) episode on the pod!

[Pedro Domingos @pmddomingos]

The biggest risk to Nvidia’s market cap is people noticing that deep learning is just scaled-up nearest neighbor.
https://t.co/irTkCbiv84