@jessicahullman.bsky.social in the comments: "Iâve had similar experiences â as soon as model fitting runs into a problem, coding agents start devising their own idiosyncratic solutions" oh no. this plus [relentless proactivity](simonwillison.net/2026/Jun/11/...) is a nightmare
Who's Who of AI
tj mahr đ€
Why they matter
Directory member with public evidence across AI research.
- AI signals
- 0 past 30d
- Sources
- 0 distinct domains
- DiscussÔes
- 33 past 30d
- Latest signal
- â
data scientist studying how kids learn to speak, dad, jump roper, bayesian, tjmahr.com, Madison, WI
What they're sharing
Claude Fable is relentlessly proactive simonwillison.net
Robust Inference and Model Criticism Using Bagged Posteriors arxiv.org
Montreal Forced Aligner and the state of speech-to-text alignment in 2026 arxiv.org
View on Bluesky â
arxiv.org
View on Bluesky â
Montreal Forced Aligner and the state of speech-to-text alignment in 2026 arxiv.org
View on Bluesky â
Articles & links
who up bagging their posteriors arxiv.org/abs/1912.07104
tj mahr đ€ reposted
Michael McAuliffe, Kaylynn Gunter, Michael Wagner, Morgan Sonderegger: Montreal Forced Aligner and the state of speech-to-text alignment in 2026 https://arxiv.org/abs/2606.18466 https://arxiv.org/pdf/2606.18466 https://arxiv.org/html/2606.18466
tj mahr đ€ reposted
Michael McAuliffe, Kaylynn Gunter, Michael Wagner, Morgan Sonderegger: Montreal Forced Aligner and the state of speech-to-text alignment in 2026 https://arxiv.org/abs/2606.18466 https://arxiv.org/pdf/2606.18466 https://arxiv.org/html/2606.18466
tj mahr đ€ reposted
Michael McAuliffe, Kaylynn Gunter, Michael Wagner, Morgan Sonderegger: Montreal Forced Aligner and the state of speech-to-text alignment in 2026 https://arxiv.org/abs/2606.18466 https://arxiv.org/pdf/2606.18466 https://arxiv.org/html/2606.18466
Their own posts
Recent commentary
I get depressed when I review something and it seems AI-assisted (as in well written but shallow and making weird jumps in spots). Why am I wasting my time putting in thought that wasnât given in the first place
feeling great about AI with my job at the âcounting letters in wordsâ factory
They keep cranking out math things at OpenAI bc itâs addicting
an old-school roomba would just cruise around until its battery was going to die, and then they tried to make it smart and map stuff and wifi-connected and they got really bad. i feel like there's an analogy here for current AI, like how it's the effort and relentlessness that's powerful
I have three instances where my role qua statistician involved being the human who had to review and vouch for the analysis technique and code provided by ChatGPT/Claude
discourse about ai has swallowed my timeline
This new verbal style. Itâs all over the place. Okay, YouTube. Terse. I canât stand it. I blame AI.
i had the llm extract part of an old C++ windows program (circa 2005) into a CLI app and then i told it to memoize some of the work (it does grapheme-to-phoneme conversion) and got a 5x performance boost
MS Teams is missing the last six months of messages in each of my chats. It's weird how AI is revolutionizing programming and yet Microsoft stuff has never seemed so janky.
just unplug the AI before it can eliminate humanity
Their network
In tj mahr đ€'s orbit
Center = tj mahr đ€. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.
Are you tj mahr đ€? Show it.
Add the Whoâs Who of AI badge to your site or bio. It links back to this profile.
Markdown: [](https://aiweekly.co/whos-who/person/tjmahr-com)