Peter
banner
competerscience.bsky.social
Peter
@competerscience.bsky.social
Postdoctoral researcher at McGill in #AI #ML core developer of SpeechBrain.

Studies analysis of speech patterns for bio markers, for speech enhancement, robust ASR, continual learning, etc.
I’m excited to see this, I feel it meaningfully moves the needle on discussions of AI and personhood. The comparison with legally holding boats responsible for damage in certain cases made things really clear for me. Great to see Pragmatism getting some of the love it deserves!
[1/9] Excited to share our new paper "A Pragmatic View of AI Personhood" published today. We feel this topic is timely, and rapidly growing in importance as AI becomes agentic, as AI agents integrate further into the economy, and as more and more users encounter AI.
November 2, 2025 at 3:13 AM
We're excited to introduce forced alignment support in SpeechBrain, thanks to Zeyu Zhao! This new feature leverages CTC-based ASR models and the power of k2, a differentiable WFST toolkit, to perform efficient and accurate forced alignment.

Check it out: speechbrain.readthedocs.io/en/develop/t...
Force alignment using k2 for CTC models. — SpeechBrain 0.5.0 documentation
speechbrain.readthedocs.io
July 18, 2025 at 2:47 AM
Fascinating analysis of the implicit values present in machine learning papers, worth pondering if you're in the field -- dl.acm.org/doi/pdf/10.1...
dl.acm.org
May 21, 2025 at 7:48 PM
Reposted by Peter
Do I know any AI engineers that have experience training LoRAs and might be looking for work? Preference for Europe/Africa time zones (DM me)
April 11, 2025 at 1:41 PM
Read my new blog post - a conversation between me and ChatGPT about disability and buddhism: peterplantinga.substack.com/p/the-disabl...
"The Disabled God" for Buddhism
I’m posting a new format this time, a chat between myself and ChatGPT.
peterplantinga.substack.com
April 6, 2025 at 3:29 PM
Reposted by Peter
I’ve long used FiveThirtyEight’s interactive “Hack Your Way To Scientific Glory” to illustrate the idea of p-hacking when I teach statistics. But ABC/Disney killed the site earlier this month :(

So I made my own with #rstats and Observable and #QuartoPub ! stats.andrewheiss.com/hack-your-way/
March 20, 2025 at 6:30 PM
I have a new substack up about the awesomeness that is #Wicked and its connection to #disability themes and #bodyqueer themes open.substack.com/pub/peterpla...
Wicked is Bodyqueer at Heart
I saw the new (2024) Wicked film adaptation and, wow what a film.
open.substack.com
January 7, 2025 at 1:58 PM
Reposted by Peter
i asked Claude to write a Barthelme-esque short story with the aesthetic sensibilities of "The School", and it gave me this. i mean. i mean.
December 16, 2024 at 4:48 AM
Reposted by Peter
Researchers reduce bias in AI models while preserving or improving accuracy https://news.mit.edu/2024/researchers-reduce-bias-ai-models-while-preserving-improving-accuracy-1211 #AI
December 14, 2024 at 12:46 AM
Proud of this work published a few days ago at SLT 2024 on continual learning for end to end ASR. Turns out changing the CL paradigm to parallel training on different tasks and merging these experts can reduce the forgetting rate to as low as 0.4%! poonehmousavi.github.io/assets/publi...
poonehmousavi.github.io
December 9, 2024 at 12:38 AM
I love Arrival so much, this makes me happy
Did you know that attention across the whole input span was inspired by the time-negating alien language in Arrival? Crazy anecdote from the latest Hard Fork podcast (by @kevinroose.com and @caseynewton.bsky.social). HT nwbrownboi on Threads for the lead.
December 1, 2024 at 2:54 PM
I’m participating in my first big #ML challenge, does anyone have some tips for me? www.drivendata.org/competitions...
PREPARE Challenge - Phase 2: Model Arena (Acoustic Track)
Advance algorithms and analytic approaches for early prediction of Alzheimer's disease and related dementias based on voice recordings, with an emphasis on explainability of predictions. [Model Arena ...
www.drivendata.org
November 29, 2024 at 2:25 PM