Cam Linke

CEO @AmiiThinks. MSc @rlai_lab

[Marlos C. Machado @MarlosCMachado]

The Department of Computing Science at the University of Alberta is hiring for two new tenure-track positions this year. Please spread the word:

- Computer Systems for AI and High-Performance Computing: https://t.co/rMJoJXwJFs

- AI for Health: https://t.co/SHBJYJdqIP

[Bahareh Tolooshams @BTolooshams]

🚨New preprint: Interference Beyond Geometry in Concept Extraction

Interference between learned features is almost always measured as geometric overlap. But geometry only tells you what could interfere. Co-activation tells you what does!

We introduce effective interference: geometry × co-activation statistics. It captures not just how features overlap, but how they are actually used together.

Why it matters for sparse autoencoders (SAEs): the encoder produces the codes, so how you build an SAE changes which interference it learns.

This is a work with Valérie Costa (@_valerie_costa_) done at @UAlberta/@AmiiThinks. Valérie is now at @Harvard/@KempnerInst for her PhD studies. 1/

paper: https://t.co/vJpR3M8KFO

[Csaba Szepesvari @CsabaSzepesvari]

https://t.co/SxXB2wiVCJ
UofA Math Dept has an opening for Tenure-Track Assistant Professor, subtitled "Mathematical Foundations of AI and Formal Reasoning". Please spread the word! Math dept is excellent, we have a lot of fun together. DM if you are interested and have questions

[Bahareh Tolooshams @BTolooshams]

We introduce Sparse Autoencoder Neural Operators (SAE-NOs), a functional framework for representation learning and mechanistic interpretability that treats data as samples from underlying continuous functions and learns mappings between function spaces.

Standard SAEs (SAE-MLP) represent each concept with a scalar activation and a vector-valued dictionary atom, limiting their ability to capture how and where a concept is expressed across structured domains.

SAE-FNO introduces feature-map representations with both concept sparsity and domain sparsity, allowing the model to capture not only which concepts are active, but also where and how they are expressed across the domain.

This is a joint collaboration, between @UAlberta/@AmiiThinks and @Caltech, with Ailsa Shen and @AnimaAnandkumar. 1/

arXiv: https://t.co/EphbL2FJYA

[Cam Linke @camlinke]

This is really great work by Erik out of @MarlosCMachado's lab. Options are an incredibly fertile direction of research in AI, with option discovery an outstanding challenge. This is a great advance for general, online, option discovery!

[Marlos C. Machado @MarlosCMachado]

I'm excited to share Erik's work!

Mastering Atari 2600 Games with Discovered Options

We have a general-purpose option-discovery method that works w/ high-dimensional observations, tabula rasa, using a single stream of experience! It outperforms DreamerV3, Rainbow, and IQN! 1/9

[Marlos C. Machado @MarlosCMachado]

I'm excited to share Erik's work!

Mastering Atari 2600 Games with Discovered Options

We have a general-purpose option-discovery method that works w/ high-dimensional observations, tabula rasa, using a single stream of experience! It outperforms DreamerV3, Rainbow, and IQN! 1/9

[Sonya Huang 🐥 @sonyatweetybird]

Today we release my favorite episode of Training Data yet: the great Rich Sutton.

@RichardSSutton wrote the textbook, wrote The Bitter Lesson (and many other on-point essays like "Self-Verification, The Key to AI"), and trained a mafia of talented students who went on to change the AI landscape forever including David Silver, inventor of built AlphaGo.

@kjaved_ was Rich's PhD student at Alberta and wrote The Big World Hypothesis. They just left academia to start @oaklab

Their core argument:
(1) The Bitter Lesson: the world is massively more complex than any model of it, so anything trained on human-curated data has a ceiling
(2) Continual Learning: intelligence is continual by definition, and today's models stop learning the moment they ship.

The conversation covers:
— what The Bitter Lesson actually says, and what people get wrong
— why synthetic data is "just a big mistake," and the Big World Hypothesis behind it
— how LLMs are both a positive and a negative example of his own essay
— why no animal learns by supervised learning, and what squirrels can do that we can't
— the cure for catastrophic forgetting: per-weight step sizes and continual backprop
— why the biggest labs can't take a path where performance gets worse before it gets better
— a trillion parameters on 20 watts, and the Moore's Law math that makes it plausible
— why the endpoint isn't one mind but one design, running as many minds

It was both a fun generative idea- and debate-filled conversation, and a surprisingly human one too. Rich, thank you for beating cancer and changing the trajectory of AI. 💙

00:00 Introduction
02:10 An AI winter, a cancer diagnosis, and the move to Alberta
07:07 Writing "The Bitter Lesson," and what people get wrong
09:53 Are LLMs a positive or a negative example of it?
11:03 Synthetic data is "just a big mistake," and the Big World Hypothesis
18:01 AlphaGo, human priors, and why prior knowledge and learning should be friends
22:37 "Their weights never change": do LLM assistants actually learn?
26:09 Babies, squirrels, and why no animal learns by supervised learning
32:02 Rockets, imagination, and where paradigm shifts come from
36:42 The Alberta Plan and its 12 steps
38:53 Catastrophic forgetting and the cure
43:43 Oak's biggest ambition: a self-maintaining mind
47:56 Why the big labs are stuck in a local minimum
49:13 If everything goes right: LLMs, many minds, and hiring

The man who pioneered reinforcement learning thinks the rest of the field is weird, and lays it all out in today's episode. Together w/ @Alfred_Lin @sequoia

[Danielle Smith @ABDanielleSmith]

Another full Stampede schedule to finish up yesterday…

Great meeting with Goldy Hyder @goldyhyder and the Business Council of Canada @BizCouncilofCan to talk about Alberta’s economic growth and the incredible opportunities ahead.

I also joined Minister RJ Sigurdson @RjSigurdson and Minister Nathan Neudorf @neudorf_ab at their Stampede Open House with stakeholders in Alberta’s utility industry.

Then it was off to Platform Calgary with Minister Nate Glubish @nateglubish, Minister Myles McDougall @McDougallYYC, Federal Minister Evan Solomon @EvanLSolomon, and Cam Linke @camlinke from the Alberta Machine Intelligence Institute @AmiiThinks to announce Alberta’s $50 million investment into AI innovation.

From AI tools helping doctors spend more time with patients, to training the next generation of Alberta workers, we are making sure our province is leading in the technologies that will shape the future.

[Evan Solomon @EvanLSolomon]

Building Canada’s AI future together!

Great to be at Platform Calgary as the Government of Alberta renewed its support for Amii, complementing our government’s investments to strengthen Canada’s national AI institutes.

We're investing in the researchers, entrepreneurs and institutions that will help ensure AI is developed, commercialized and scaled right here in Canada.

🇨🇦
Nous bâtissons ensemble l'avenir de l'IA au Canada!

Ravi d'être à Platform Calgary alors que le gouvernement de l'Alberta renouvelle son soutien à Amii, en complément des investissements de notre gouvernement, qui visent à renforcer les instituts nationaux d'IA du Canada.

Nous investissons dans les chercheurs, les entrepreneurs et les institutions qui contribueront à faire en sorte que l'IA soit développée, commercialisée et déployée à grande échelle ici même, au Canada.

[Peter Tertzakian @PTertzakian]

At @AmiiThinks’ #UpperBound2026, I sat down with @camlinke to discuss the relationship between AI and energy (spoiler: AI is 100% an energy story), the geopolitical realities of this modern computing arms race, and how AI can recursively modernize the energy sector.
https://t.co/VSzezpIYjd

[Evan Solomon @EvanLSolomon]

AI for All.

🇨🇦

L’IA pour tous.

[Evan Solomon @EvanLSolomon]

Great to be there with you and the incredible builders at Upper Bound on Edmonton. Over 11000 attendees and great to announce the new and returning CIFAR Chairs for more research and innovation on Canada. Thanks for your leadership Cam.

[Cam Linke @camlinke]

Great to host @EvanLSolomon in Edmonton this week at Upper Bound to talk about Canada's AI leadership!

[Evan Solomon @EvanLSolomon]

This morning at the Upper Bound AI Conference in Edmonton, I had a great conversation with @camlinke on where AI is heading and what it will take for Canada to lead in this new economy.

We talked about the importance of building trusted AI infrastructure, supporting Canadian talent and giving companies the tools they need to grow, compete and create good jobs here at home. Canada has the expertise and ambition to help shape the future of AI.
🇨🇦
Ce matin, lors de la conférence d'IA Upper Bound à Edmonton, j’ai eu une conversation très intéressante avec @camlinke sur l’avenir de l’IA et sur ce qu’il faudra au Canada pour devenir un leader dans cette nouvelle économie.

Nous avons discuté de l’importance de mettre en place une infrastructure d’IA fiable, de soutenir les talents canadiens et de donner aux entreprises les outils dont elles ont besoin pour se développer, être compétitives et créer des emplois de qualité ici, chez nous. Le Canada possède l’expertise et l’ambition nécessaires pour contribuer à façonner l’avenir de l’IA.

[Evan Solomon @EvanLSolomon]

This morning at the Upper Bound AI Conference in Edmonton, I had a great conversation with @camlinke on where AI is heading and what it will take for Canada to lead in this new economy.

We talked about the importance of building trusted AI infrastructure, supporting Canadian talent and giving companies the tools they need to grow, compete and create good jobs here at home. Canada has the expertise and ambition to help shape the future of AI.
🇨🇦
Ce matin, lors de la conférence d'IA Upper Bound à Edmonton, j’ai eu une conversation très intéressante avec @camlinke sur l’avenir de l’IA et sur ce qu’il faudra au Canada pour devenir un leader dans cette nouvelle économie.

Nous avons discuté de l’importance de mettre en place une infrastructure d’IA fiable, de soutenir les talents canadiens et de donner aux entreprises les outils dont elles ont besoin pour se développer, être compétitives et créer des emplois de qualité ici, chez nous. Le Canada possède l’expertise et l’ambition nécessaires pour contribuer à façonner l’avenir de l’IA.

[Cam Linke @camlinke]

Whatever streaming service puts together @Jaboowins and @qb_fernando for the most wholesome show of all time will instantly get my subscription. That show would single handedly revive the motivational poster industry.

[Erin Grant @ermgrant]

Thrilled to announce I'll start in 2026 as faculty in Psych & CS @UAlberta + @AmiiThinks Fellow!! 🥳 Recruiting students to develop theories of cognition in natural and artificial systems 🤖💭🧠. Find me at #NeurIPS2025 workshops (talk at @CogInterp & organising @DataOnBrainMind)

[Richard Sutton @RichardSSutton]

@eigenrobot Even in birdsong learning in zebra finches the motor actions are not learned by imitation. The auditory result is reproduced, not the actions; in this crucial way it differs from LLM training.

[Evan Solomon @EvanLSolomon]

Today in Edmonton, we announced new federal support to strengthen Canada’s AI compute infrastructure—giving researchers and innovators the tools they need to drive the next great discovery.

We also announced new support to help Canadian workers gain the skills they need for the jobs of the future.

Thanks to @camlinke, the team at @AmiiThinks, and @EleanorO_MP for hosting @PattyHajdu and me.

[Amii @AmiiThinks]

We were pleased to welcome Minister @EvanLSolomon to Amii HQ today. Amii CEO @CamLinke joined Minister Solomon for a tour and a great fireside chat about #AI in Canada. 🇨🇦

Stay tuned for a video and recap of that conversation!

[Cam Linke @camlinke]

Great to have Canada's AI Minister @EvanLSolomon visit us at @AmiiThinks today to talk about Canada's leadership in AI!

[Evan Solomon @EvanLSolomon]

Currently at @AmiiThinks in Edmonton with CEO @camlinke and AI pioneer @RichardSSutton.

One of Canada’s 3 national AI institutes, Amii is turning world-class research into real-world results—training talent, backing start-ups, and helping local businesses put AI to work.

Amii proves that innovation isn’t just happening in Silicon Valley—it’s happening right here in Edmonton.