What is the charge? Spending on inference? A succulent Chinese inference??
November 6, 2026 at 9:41 PM
What is the charge? Spending on inference? A succulent Chinese inference??
Causal inference is my passion.
November 5, 2026 at 3:42 PM
Causal inference is my passion.
hot take: yes please ban all datacenter construction. That will force labs to make inference cheap sooner rather than later
November 7, 2026 at 1:25 PM
hot take: yes please ban all datacenter construction. That will force labs to make inference cheap sooner rather than later
file under: context, presupposition, inference
November 7, 2026 at 4:06 AM
file under: context, presupposition, inference
Kids: study causal inference. I wouldn't accept this from an undergraduate in my intro stats class.
November 3, 2026 at 10:07 PM
Kids: study causal inference. I wouldn't accept this from an undergraduate in my intro stats class.
Universal basic inference (:
November 5, 2026 at 9:17 AM
Universal basic inference (:
Nice inference!
November 5, 2026 at 10:47 PM
Nice inference!
Apropos of nothing, this is a regular PSA that the excellent book “Computer Age Statistical Inference” is available free online.
hastie.su.domains/CASI/
hastie.su.domains/CASI/
November 5, 2026 at 5:30 PM
Apropos of nothing, this is a regular PSA that the excellent book “Computer Age Statistical Inference” is available free online.
hastie.su.domains/CASI/
hastie.su.domains/CASI/
Ecological inference for now but suggestive www.economist.com/interactive/...
November 4, 2026 at 2:07 PM
Ecological inference for now but suggestive www.economist.com/interactive/...
A bit of a milestone!
November 6, 2026 at 9:50 PM
A bit of a milestone!
if we assume he will win his seat next cycle, is it bayhesian inference?
November 5, 2026 at 6:18 PM
if we assume he will win his seat next cycle, is it bayhesian inference?
“He’s probably fine with them.” Another racial inference from a lifelong racist. It generationally ingrained and he just cannot help himself. I fucking hate him.
November 5, 2026 at 9:47 PM
“He’s probably fine with them.” Another racial inference from a lifelong racist. It generationally ingrained and he just cannot help himself. I fucking hate him.
I always love the inference that I invented it. “howd you think of that?”
November 7, 2026 at 7:37 PM
I always love the inference that I invented it. “howd you think of that?”
I̶n̶f̶e̶r̶e̶n̶c̶e̶ ̶I̶n̶n̶s̶
G̶i̶g̶a̶w̶a̶t̶t̶ ̶G̶a̶r̶d̶e̶n̶s̶
S̶i̶l̶i̶c̶o̶n̶ ̶S̶a̶n̶c̶t̶u̶a̶r̶i̶e̶s̶
B̶r̶a̶i̶n̶ ̶B̶a̶r̶n̶s̶
I̶n̶f̶e̶r̶e̶n̶c̶e̶ ̶I̶g̶l̶o̶o̶s̶
A̶l̶g̶o̶r̶i̶t̶h̶m̶ ̶A̶r̶b̶o̶r̶e̶t̶u̶m̶s̶
C̶o̶n̶s̶c̶i̶o̶u̶s̶n̶e̶s̶s̶ ̶C̶o̶n̶s̶e̶r̶v̶a̶t̶o̶r̶i̶e̶s̶
Compute Campuses
G̶i̶g̶a̶w̶a̶t̶t̶ ̶G̶a̶r̶d̶e̶n̶s̶
S̶i̶l̶i̶c̶o̶n̶ ̶S̶a̶n̶c̶t̶u̶a̶r̶i̶e̶s̶
B̶r̶a̶i̶n̶ ̶B̶a̶r̶n̶s̶
I̶n̶f̶e̶r̶e̶n̶c̶e̶ ̶I̶g̶l̶o̶o̶s̶
A̶l̶g̶o̶r̶i̶t̶h̶m̶ ̶A̶r̶b̶o̶r̶e̶t̶u̶m̶s̶
C̶o̶n̶s̶c̶i̶o̶u̶s̶n̶e̶s̶s̶ ̶C̶o̶n̶s̶e̶r̶v̶a̶t̶o̶r̶i̶e̶s̶
Compute Campuses
OpenAI might rebrand its data centers as ‘AI compute campuses’
OpenAI might rebrand its data centers as ‘AI compute campuses’
Data centers are deeply unpopular. But what if we called them something else?
techcrunch.com
November 2, 2026 at 2:43 AM
I̶n̶f̶e̶r̶e̶n̶c̶e̶ ̶I̶n̶n̶s̶
G̶i̶g̶a̶w̶a̶t̶t̶ ̶G̶a̶r̶d̶e̶n̶s̶
S̶i̶l̶i̶c̶o̶n̶ ̶S̶a̶n̶c̶t̶u̶a̶r̶i̶e̶s̶
B̶r̶a̶i̶n̶ ̶B̶a̶r̶n̶s̶
I̶n̶f̶e̶r̶e̶n̶c̶e̶ ̶I̶g̶l̶o̶o̶s̶
A̶l̶g̶o̶r̶i̶t̶h̶m̶ ̶A̶r̶b̶o̶r̶e̶t̶u̶m̶s̶
C̶o̶n̶s̶c̶i̶o̶u̶s̶n̶e̶s̶s̶ ̶C̶o̶n̶s̶e̶r̶v̶a̶t̶o̶r̶i̶e̶s̶
Compute Campuses
G̶i̶g̶a̶w̶a̶t̶t̶ ̶G̶a̶r̶d̶e̶n̶s̶
S̶i̶l̶i̶c̶o̶n̶ ̶S̶a̶n̶c̶t̶u̶a̶r̶i̶e̶s̶
B̶r̶a̶i̶n̶ ̶B̶a̶r̶n̶s̶
I̶n̶f̶e̶r̶e̶n̶c̶e̶ ̶I̶g̶l̶o̶o̶s̶
A̶l̶g̶o̶r̶i̶t̶h̶m̶ ̶A̶r̶b̶o̶r̶e̶t̶u̶m̶s̶
C̶o̶n̶s̶c̶i̶o̶u̶s̶n̶e̶s̶s̶ ̶C̶o̶n̶s̶e̶r̶v̶a̶t̶o̶r̶i̶e̶s̶
Compute Campuses
i love how horribly inefficient / uneconomical running at-home inference is at full compute capacity lmao it's ridiculous. i still want to do it.
November 5, 2026 at 5:10 PM
i love how horribly inefficient / uneconomical running at-home inference is at full compute capacity lmao it's ridiculous. i still want to do it.
*free use as a model inference/fine-tune provider
November 4, 2026 at 4:36 PM
*free use as a model inference/fine-tune provider
good news: switching to GPT-6 for my inference is faster, cheaper, and quieter
bad news: oh my god it writes the ugliest code what have they done to these models
bad news: oh my god it writes the ugliest code what have they done to these models
November 6, 2026 at 6:09 PM
good news: switching to GPT-6 for my inference is faster, cheaper, and quieter
bad news: oh my god it writes the ugliest code what have they done to these models
bad news: oh my god it writes the ugliest code what have they done to these models
This is an important book that a lot of people need to read!
www.cambridge.org/core/books/q...
www.cambridge.org/core/books/q...
Qualitative Causal Inference and Explanation
Cambridge Core - Qualitative Research Methods - Qualitative Causal Inference and Explanation
www.cambridge.org
November 6, 2026 at 7:35 PM
This is an important book that a lot of people need to read!
www.cambridge.org/core/books/q...
www.cambridge.org/core/books/q...
This is causal inference
November 6, 2026 at 12:23 AM
This is causal inference
Oh one other thing.
The stable inference margins for Q3 vs 1H suggests to me that the SPCX compute is not being used for inference, but instead for training (capex).
That makes it uh...incredibly vulnerable if they decide to brake development.
The stable inference margins for Q3 vs 1H suggests to me that the SPCX compute is not being used for inference, but instead for training (capex).
That makes it uh...incredibly vulnerable if they decide to brake development.
November 2, 2026 at 3:04 PM
Oh one other thing.
The stable inference margins for Q3 vs 1H suggests to me that the SPCX compute is not being used for inference, but instead for training (capex).
That makes it uh...incredibly vulnerable if they decide to brake development.
The stable inference margins for Q3 vs 1H suggests to me that the SPCX compute is not being used for inference, but instead for training (capex).
That makes it uh...incredibly vulnerable if they decide to brake development.
inference prices instead
November 4, 2026 at 3:21 PM
inference prices instead
Bayesian phylogenetic inference accounting for recombination in whole bacterial genomes https://www.biorxiv.org/content/10.64898/2026.11.04.764181v1
November 5, 2026 at 1:17 PM
Bayesian phylogenetic inference accounting for recombination in whole bacterial genomes https://www.biorxiv.org/content/10.64898/2026.11.04.764181v1
#vLLM control plane in #k8s for inference managment with #kv-cache, policies, routing, and load balancing
#OpenSource Apache 2.0
#AI #LLM #Kubernetes #NVIDIA
#OpenSource Apache 2.0
#AI #LLM #Kubernetes #NVIDIA
GitHub - Kinference-AI/kinference: An inference control plane for Kubernetes and vLLM
An inference control plane for Kubernetes and vLLM - Kinference-AI/kinference
github.com
November 5, 2026 at 5:48 PM
#vLLM control plane in #k8s for inference managment with #kv-cache, policies, routing, and load balancing
#OpenSource Apache 2.0
#AI #LLM #Kubernetes #NVIDIA
#OpenSource Apache 2.0
#AI #LLM #Kubernetes #NVIDIA
Me, watching the US AI industry get its comeuppance while 90% of my inference is now French hosted open models.
November 6, 2026 at 8:15 PM
Me, watching the US AI industry get its comeuppance while 90% of my inference is now French hosted open models.
Remarkably, despite those higher compute costs, inference margins were 85.4% in Q3 (we know because we can back out the previously disclosed 1H numbers from then new 9m '26 numbers). Again, they're printing money in inference despite the high marginal compute cost.
November 2, 2026 at 2:48 PM
Remarkably, despite those higher compute costs, inference margins were 85.4% in Q3 (we know because we can back out the previously disclosed 1H numbers from then new 9m '26 numbers). Again, they're printing money in inference despite the high marginal compute cost.