Warren
@brokenglobe.bsky.social
Artist and Technologist
previously #yeg, now #hnd
previously #yeg, now #hnd
I am hearing this directly from my Chinese partners, there are (or were) checks and balances to some degree in Chinese leadership, Xi is not a "king" ... or was not ... it's worrisome even to insiders.
February 3, 2026 at 9:16 AM
I am hearing this directly from my Chinese partners, there are (or were) checks and balances to some degree in Chinese leadership, Xi is not a "king" ... or was not ... it's worrisome even to insiders.
Ad hominem but Nutlick is as reliable as Baghdad Bob.
January 20, 2026 at 10:34 PM
Ad hominem but Nutlick is as reliable as Baghdad Bob.
I think he's trying to be hopeful recognizing that there is no path to prevent or undo what Trump is doing.
January 20, 2026 at 10:32 PM
I think he's trying to be hopeful recognizing that there is no path to prevent or undo what Trump is doing.
6/6
The days of relying on the cloud for every smart interaction are numbered. FunctionGemma signals a shift toward compact, efficient, and private AI agents that live right in your pocket.
Asif Razzaq | Marktechpost
www.marktechpost.com/2025/12/26/f...
The days of relying on the cloud for every smart interaction are numbered. FunctionGemma signals a shift toward compact, efficient, and private AI agents that live right in your pocket.
Asif Razzaq | Marktechpost
www.marktechpost.com/2025/12/26/f...
From Gemma 3 270M to FunctionGemma, How Google AI Built a Compact Function Calling Specialist for Edge Workloads
From Gemma 3 270M to FunctionGemma, How Google AI Built a Compact Function Calling Specialist for Edge Workloads
www.marktechpost.com
December 31, 2025 at 7:13 PM
6/6
The days of relying on the cloud for every smart interaction are numbered. FunctionGemma signals a shift toward compact, efficient, and private AI agents that live right in your pocket.
Asif Razzaq | Marktechpost
www.marktechpost.com/2025/12/26/f...
The days of relying on the cloud for every smart interaction are numbered. FunctionGemma signals a shift toward compact, efficient, and private AI agents that live right in your pocket.
Asif Razzaq | Marktechpost
www.marktechpost.com/2025/12/26/f...
5/6
Google is proving it with "Tiny Garden," where voice commands plant virtual sunflowers, and a physics playground solved by natural language.
It’s multi-step logic handled entirely on-device, validating a future where privacy and performance coexist.
Google is proving it with "Tiny Garden," where voice commands plant virtual sunflowers, and a physics playground solved by natural language.
It’s multi-step logic handled entirely on-device, validating a future where privacy and performance coexist.
December 31, 2025 at 7:13 PM
5/6
Google is proving it with "Tiny Garden," where voice commands plant virtual sunflowers, and a physics playground solved by natural language.
It’s multi-step logic handled entirely on-device, validating a future where privacy and performance coexist.
Google is proving it with "Tiny Garden," where voice commands plant virtual sunflowers, and a physics playground solved by natural language.
It’s multi-step logic handled entirely on-device, validating a future where privacy and performance coexist.
4/6
Trained on 6 trillion tokens, it learns both syntax and intent.
Out of the box, it hits 58% accuracy on mobile actions. But give it a little fine-tuning? That reliability jumps to 85%, proving small models can punch way above their weight class.
Trained on 6 trillion tokens, it learns both syntax and intent.
Out of the box, it hits 58% accuracy on mobile actions. But give it a little fine-tuning? That reliability jumps to 85%, proving small models can punch way above their weight class.
December 31, 2025 at 7:12 PM
4/6
Trained on 6 trillion tokens, it learns both syntax and intent.
Out of the box, it hits 58% accuracy on mobile actions. But give it a little fine-tuning? That reliability jumps to 85%, proving small models can punch way above their weight class.
Trained on 6 trillion tokens, it learns both syntax and intent.
Out of the box, it hits 58% accuracy on mobile actions. But give it a little fine-tuning? That reliability jumps to 85%, proving small models can punch way above their weight class.
3/6
Designed for the "edge"—think laptops and phones rather than massive server farms.
With a 32K token context window and optimization for JSON structures, it runs locally where memory is tight and speed is non-negotiable.
Designed for the "edge"—think laptops and phones rather than massive server farms.
With a 32K token context window and optimization for JSON structures, it runs locally where memory is tight and speed is non-negotiable.
December 31, 2025 at 7:11 PM
3/6
Designed for the "edge"—think laptops and phones rather than massive server farms.
With a 32K token context window and optimization for JSON structures, it runs locally where memory is tight and speed is non-negotiable.
Designed for the "edge"—think laptops and phones rather than massive server farms.
With a 32K token context window and optimization for JSON structures, it runs locally where memory is tight and speed is non-negotiable.
2/6
It’s a 270M parameter powerhouse built on the Gemma 3 architecture. Unlike chatty assistants, this specialist doesn't just talk—it *does*.
It translates your words into structured function calls, bridging the gap between human intent and software execution.
It’s a 270M parameter powerhouse built on the Gemma 3 architecture. Unlike chatty assistants, this specialist doesn't just talk—it *does*.
It translates your words into structured function calls, bridging the gap between human intent and software execution.
December 31, 2025 at 7:11 PM
2/6
It’s a 270M parameter powerhouse built on the Gemma 3 architecture. Unlike chatty assistants, this specialist doesn't just talk—it *does*.
It translates your words into structured function calls, bridging the gap between human intent and software execution.
It’s a 270M parameter powerhouse built on the Gemma 3 architecture. Unlike chatty assistants, this specialist doesn't just talk—it *does*.
It translates your words into structured function calls, bridging the gap between human intent and software execution.
8/8
The era of ad-free AI assistance appears to be closing. As the platform evolves from a tool into a marketplace, that "trusted relationship" faces its biggest test yet.
Joe Wilkins | Futurism
futurism.com/artificial-i...
The era of ad-free AI assistance appears to be closing. As the platform evolves from a tool into a marketplace, that "trusted relationship" faces its biggest test yet.
Joe Wilkins | Futurism
futurism.com/artificial-i...
OpenAI Reportedly Planning to Make ChatGPT "Prioritize" Advertisers in Conversation
OpenAI employees working on ChatGPT report plans to unleash sponsored advertisements above organic results.
futurism.com
December 31, 2025 at 7:03 PM
8/8
The era of ad-free AI assistance appears to be closing. As the platform evolves from a tool into a marketplace, that "trusted relationship" faces its biggest test yet.
Joe Wilkins | Futurism
futurism.com/artificial-i...
The era of ad-free AI assistance appears to be closing. As the platform evolves from a tool into a marketplace, that "trusted relationship" faces its biggest test yet.
Joe Wilkins | Futurism
futurism.com/artificial-i...
7/8
OpenAI claims this is about "offering more intelligence to everyone."
A spokesperson insists that because people have a "trusted relationship" with ChatGPT, any ad integration would be designed to somehow "respect that trust."
OpenAI claims this is about "offering more intelligence to everyone."
A spokesperson insists that because people have a "trusted relationship" with ChatGPT, any ad integration would be designed to somehow "respect that trust."
December 31, 2025 at 7:02 PM
7/8
OpenAI claims this is about "offering more intelligence to everyone."
A spokesperson insists that because people have a "trusted relationship" with ChatGPT, any ad integration would be designed to somehow "respect that trust."
OpenAI claims this is about "offering more intelligence to everyone."
A spokesperson insists that because people have a "trusted relationship" with ChatGPT, any ad integration would be designed to somehow "respect that trust."
6/8
This isn't just talk. The move to cram commercials into the interface was confirmed when code surfaced in the Android beta app.
Sleuths uncovered explicit references to "feature ads" and "search ads carousel" buried within the software lines.
This isn't just talk. The move to cram commercials into the interface was confirmed when code surfaced in the Android beta app.
Sleuths uncovered explicit references to "feature ads" and "search ads carousel" buried within the software lines.
December 31, 2025 at 7:02 PM
6/8
This isn't just talk. The move to cram commercials into the interface was confirmed when code surfaced in the Android beta app.
Sleuths uncovered explicit references to "feature ads" and "search ads carousel" buried within the software lines.
This isn't just talk. The move to cram commercials into the interface was confirmed when code surfaced in the Android beta app.
Sleuths uncovered explicit references to "feature ads" and "search ads carousel" buried within the software lines.
5/8
To avoid putting users off entirely, internal mockups suggest a slower rollout.
One proposal shows ads appearing only after the second prompt—a calculated attempt to hook you with utility before "bombarding" you with commercial content.
To avoid putting users off entirely, internal mockups suggest a slower rollout.
One proposal shows ads appearing only after the second prompt—a calculated attempt to hook you with utility before "bombarding" you with commercial content.
December 31, 2025 at 7:02 PM
5/8
To avoid putting users off entirely, internal mockups suggest a slower rollout.
One proposal shows ads appearing only after the second prompt—a calculated attempt to hook you with utility before "bombarding" you with commercial content.
To avoid putting users off entirely, internal mockups suggest a slower rollout.
One proposal shows ads appearing only after the second prompt—a calculated attempt to hook you with utility before "bombarding" you with commercial content.
4/8
With a self-reported 900 million weekly users, the consequences of this shift are staggering.
OpenAI isn't blind to the risk. They know corporate advertisers looking on like a "third-wheel" could kill the vibe that drives hours-long chat sessions.
With a self-reported 900 million weekly users, the consequences of this shift are staggering.
OpenAI isn't blind to the risk. They know corporate advertisers looking on like a "third-wheel" could kill the vibe that drives hours-long chat sessions.
December 31, 2025 at 7:01 PM
4/8
With a self-reported 900 million weekly users, the consequences of this shift are staggering.
OpenAI isn't blind to the risk. They know corporate advertisers looking on like a "third-wheel" could kill the vibe that drives hours-long chat sessions.
With a self-reported 900 million weekly users, the consequences of this shift are staggering.
OpenAI isn't blind to the risk. They know corporate advertisers looking on like a "third-wheel" could kill the vibe that drives hours-long chat sessions.
3/8
Imagine asking for a headache cure and getting a promoted Advil ad before the dosage instructions.
Employees fear legitimate health data could be buried under a mountain of ad text, forcing users to scroll past sponsors just to find the truth.
Imagine asking for a headache cure and getting a promoted Advil ad before the dosage instructions.
Employees fear legitimate health data could be buried under a mountain of ad text, forcing users to scroll past sponsors just to find the truth.
December 31, 2025 at 7:00 PM
3/8
Imagine asking for a headache cure and getting a promoted Advil ad before the dosage instructions.
Employees fear legitimate health data could be buried under a mountain of ad text, forcing users to scroll past sponsors just to find the truth.
Imagine asking for a headache cure and getting a promoted Advil ad before the dosage instructions.
Employees fear legitimate health data could be buried under a mountain of ad text, forcing users to scroll past sponsors just to find the truth.
2/8
Internal conversations reveal a push to unleash sponsored content that gets "preferential treatment."
The goal? Supplying advertisers with a massive audience to manipulate, potentially brushing actual answers aside for paid placements.
Internal conversations reveal a push to unleash sponsored content that gets "preferential treatment."
The goal? Supplying advertisers with a massive audience to manipulate, potentially brushing actual answers aside for paid placements.
December 31, 2025 at 6:59 PM
2/8
Internal conversations reveal a push to unleash sponsored content that gets "preferential treatment."
The goal? Supplying advertisers with a massive audience to manipulate, potentially brushing actual answers aside for paid placements.
Internal conversations reveal a push to unleash sponsored content that gets "preferential treatment."
The goal? Supplying advertisers with a massive audience to manipulate, potentially brushing actual answers aside for paid placements.
Reposted by Warren
31/31 CONCLUSION: This exhaustive analysis proves Alberta's laws are "not a good-faith effort to protect children but an ideologically motivated campaign" inflicting measurable harm on vulnerable youth.
By RichardJR: "Legislating Harm" Analysis
docs.google.com/document/d/e...
By RichardJR: "Legislating Harm" Analysis
docs.google.com/document/d/e...
Legislating Harm: An Analysis of the Impact of Alberta's Gender-Related Policies on Transgender Youth and Society
docs.google.com
July 17, 2025 at 8:08 PM
31/31 CONCLUSION: This exhaustive analysis proves Alberta's laws are "not a good-faith effort to protect children but an ideologically motivated campaign" inflicting measurable harm on vulnerable youth.
By RichardJR: "Legislating Harm" Analysis
docs.google.com/document/d/e...
By RichardJR: "Legislating Harm" Analysis
docs.google.com/document/d/e...
How about just getting the F out of Ukraine and going home? Russia has enough land and problems and opportunities without encroaching on the sovereign territory of others.
Lavrov is culpable for being a mouthpiece to the wanton murder.
Lavrov is culpable for being a mouthpiece to the wanton murder.
July 11, 2025 at 4:56 PM
How about just getting the F out of Ukraine and going home? Russia has enough land and problems and opportunities without encroaching on the sovereign territory of others.
Lavrov is culpable for being a mouthpiece to the wanton murder.
Lavrov is culpable for being a mouthpiece to the wanton murder.
Good to hear some positive feedback. I've waited more than 45 days after purchase without a shipping notification so I wasn't sure if they were a scam.
July 11, 2025 at 4:51 PM
Good to hear some positive feedback. I've waited more than 45 days after purchase without a shipping notification so I wasn't sure if they were a scam.
And what about all the illegal guns coming into Canada from the USA! They should solve our gun problems before crying over spilt (or protected) milk (and other dairy products).
Don't have a cow man.
Don't have a cow man.
July 11, 2025 at 2:55 PM
And what about all the illegal guns coming into Canada from the USA! They should solve our gun problems before crying over spilt (or protected) milk (and other dairy products).
Don't have a cow man.
Don't have a cow man.