Associate Professor, Computer Science, Texas State University. Research: Machine Learning, Time Series, Biosignal Analysis. {Follow your curiosity!}
Austin, TX
Joined May 2009
- Tweets448
- Following460
- Followers108
- Likes518
"To be a scientist is the ultimate act of charity."
Love this framing from @Ada_Palmer on the @dwarkesh_sp podcast. She’s paraphrasing Francis Bacon’s radical 17th-century idea: that the true purpose of science isn't just "truth," but the active relief of human suffering.
Hot take: we will not get true AGI until certain AI predictions are governed by Mathematics and not by Statistics.
Effective and precise communication is the new best programming language.
youtu.be/8rABwKRsec4?si=UhMk…
The implications of this paper seem big. There is something inherently problematic in how modern neural networks learn and form internal representations.
arxiv.org/abs/2505.11581
What does humanity need more right now, a super-intelligent AI or a super-wise AI? It seems to me we are in a more dire need for the latter. Why is nobody talking about it?
Looking forward to the event! Our team will present our poster titled "A Framework for End-to-End Machine Translation of American Sign Language to English for Improved Telehealth Services for the Hearing Impaired."
Health Scholar Showcase is just around the corner! Don’t miss your chance to join us for an afternoon of engaging research at the largest #txst faculty research showcase.
RSVP here and secure your spot for the 2025 #thrcHSS: bit.ly/HSSRSVP
ALT Graphic flyer titled: Health Scholar Showcase April 4th 2-5pm TXST Performing Arts Center; RSVP Today!
Vangelis Metsis retweeted
this is one of the best essays I’ve read all year and really cleanly articulates all of the thoughts I’ve been yelling to ppl about for a while
1/7 🧵 Gradient Descent vs Natural Selection: Surprising Similarities and Key Differences
#MachineLearning #Evolution #DataScience
6/7 Another Key Distinction:
• Objective Stability: The "fitness function" in NS can change with shifting environments
• GD typically optimizes a fixed objective function
LLMs & deep learning models rely on statistics, ignoring causality. This leads to probabilistic predictions with risks of errors & hallucinations. Causal reasoning enables deterministic predictions. Next-gen AI must blend statistical & causal approaches for more reliable results.
Future AI may blend probabilistic models, causal symbolic reasoning, and fact-checked retrieval for reliability. Auto-regression with selective forgetting, mimicking human working memory, could be key. This would allow discarding incorrect info and making better decisions.