Max Bain (@maxhbain) header
Max Bain (@maxhbain) profile picture

research scientist, gemini, pretraining @googledeepmind it aint that deep

Joined April 2021
Photo by Max Bain (@maxhbain)Photo by Max Bain (@maxhbain)Photo by Max Bain (@maxhbain)Photo by Max Bain (@maxhbain)Photo by Max Bain (@maxhbain)Photo by Max Bain (@maxhbain)Photo by Max Bain (@maxhbain)Photo by Max Bain (@maxhbain)Photo by Max Bain (@maxhbain)Photo by Max Bain (@maxhbain)
WhisperX version 2.0 out, now with speaker diarization and character-level timestamps. github.com/m-bain/whisperX 🧵
25
192
15
1,015
180,079
We're hiring in London and California! Fun team to be in, too! Consider applying if you would like working across the pre-training stack!
The Gemini Large Scale Pretrain team is hiring a Research Engineer in London or MTV! The team works across the pretrain stack: research codebases, distributed training infra, data pipelines, tooling for training dynamics analysis, and most importantly using Gemini to accelerate our work! 🚀 If you're an engineer who has a strong "whatever it takes" attitude and is not afraid of spending tokens, apply here with some showcase work: google.com/about/careers/app… No particular qualifications needed, and we welcome early career applicants. Join if you want me to bug you with refactors to enable [new AGI compute multiplier]! 👀
3
16
1
221
17,997
🚨 WhisperX Upgrades, Just added CPU speed up and HF fallbacks on alignment. I'm Tired of running manual checks on every PR, so I'm building an automated RSI verification loop on WhisperX. I'm starting with latency improvement PRs, but it will be getting more advanced.
2
3
4
139
New WhisperX paper dropped🚨 (yes ik, 3 years later, people are still using it somehow) Problem: WhisperX batch transcribes audio chunks, but each chunk is isolated leading to inconsistent punctuation + proper nouns⚠️ We solved this✅, WER📉. Proper noun+punctuation improves🏆
11
22
1
216
10,130
My younger bro took the baton @Carlos_O_B1, doing all the technical work. He got into ML recently, formerly an athlete, and looking for roles as a Research Engineer, if you are hiring reach out. Buy the stock while its low.
1
11
818
We have started our most ambitious pre-training run yet, for Gemini 4, and are excited by the progress : )
1,033
703
479
15,109
2,216,240
🎬 SLoMO is back! Ready to push the frontier of story-level movie understanding and Audio Description? Join us at the 2nd Story-Level Movie Understanding & Audio Description (SLoMO) Workshop at #ECCV2026 in Malmö, Sweden! 📅 Afternoon, Sept. 8, 2026 🌐 Website: slomo-workshop.github.io/ecc… 🏆 Competition details below 👇 MovieQA track: eval.ai/web/challenges/chall… Audio Description track: eval.ai/web/challenges/chall…
3
8
4
15
3,756
🗣️📣does anyone have the TEDLIUM-3 benchmark downloaded they can send my way? Seems its been taken down. Got some whisperx upgrades (with @Carlos_O_B1) and need to run evals
1
1
5
1,544
This chap is really quite misinformed
This is my fifth conversation with @GavinSBaker. Gavin understands semiconductors and AI as well as anyone I know and has a gift for making sense of the industry's complexity and nuance. We discuss: - Nvidia vs Google (GPUs + TPUs) - Scaling laws and reasoning models - The economics of AI compute - Why Blackwell's delay mattered - The bear case on the AI capex buildout - Data centers in space - The mistake SaaS companies are making Few people love investing more than Gavin. His closing answer about why he loves it turned into a full reflection on his investing origin story, which I had never heard before. Enjoy! Timestamps: 0:00 Intro 5:03 The Blackwell Transition 23:15 The Prisoner's Dilemma 27:12 The Bear Case: Edge AI 37:19 Meta, Open Source, and Model Depreciation 43:08 Geopolitics and Rare Earths 50:42 Data Centers in Space 56:06 Power Constraints as a Governor 1:11:31 The SaaS Mistake 1:16:17 Nuclear and Quantum 1:22:25 Gavin’s Investing Origins
1
10
1,714
Can Chinese LLMs beat the market? Tested 3 models on 20 volatile small-cap stocks" 🧵
1
4
1
7
886
The secret behind Gemini 3? Simple: Improving pre-training & post-training 🤯 Pre-training: Contra the popular belief that scaling is over—which we discussed in our NeurIPS '25 talk with @ilyasut and @quocleix—the team delivered a drastic jump. The delta between 2.5 and 3.0 is as big as we've ever seen. No walls in sight! Post-training: Still a total greenfield. There's lots of room for algorithmic progress and improvement, and 3.0 hasn't been an exception, thanks to our stellar team. Congratulations to the whole team 💙💙💙
117
536
227
4,322
2,031,685
With the right context LLMs can beat most day traders. I pitted 6 AI models against each other to trade a selection of S&P stocks — with access to news, earnings, sentiment etc. — over 3 months. They all outperformed s&p, with the best one (Claude) by almost 4x.
4
6
1
12
1,072
Today we're publishing an Open Letter from over 150+ founders & investors across our startup & scaleup ecosystem to the @RachelReevesMP warning against a so-called 'Exit Tax'. You can read the letter and sign on at no-exit-tax.co.uk
2
22
22
114
116,523
My lil bro @Carlos_O_B1 been in the lab, coding non stop. Watch this space.
7
416
ahola, come join the workshop tomorrow🫶🌺
The SLoMo workshop on "Story-level Movie Understanding & Audio Description" will be on #ICCV2025 day-1 morning, starting at 8:40 AM at Room 327! @JunyuXieArthur, @maxhbain and Xi will be there in person. See you tomorrow @ICCVConference !! #iccv25
15
1,313
We are hiring! I'm looking for a Research Scientist to work on Gemini Audio in @GoogleDeepMind London ♊️🔊 Feel free to reach out if you have any questions! job-boards.greenhouse.io/dee…
1
1
21
2,175
the modality gap gonna be solved in < 1 year
17
1,214
method #2 to prompt a breakthrough: run those "what if i just..." / "will it break if ..." experiments, purely out of curiosity. the results may surprise you, and give you the clue you need
why does the breakthrough always happen 0-1 days before the deadline
3
11
1,642