noble
@nobles.page
I focus on the responsible delivery of modern tech.
I did Product Mgmt but switched to Eng Mgmt
I spoke here: sessionize.com/nobleackerson
I write here: medium.com/@nobleackerson
I teach here: Udacity Agentic AI Fluency
I did Product Mgmt but switched to Eng Mgmt
I spoke here: sessionize.com/nobleackerson
I write here: medium.com/@nobleackerson
I teach here: Udacity Agentic AI Fluency
Adding the link helps:
medium.com/google-cloud...
medium.com/google-cloud...
Introducing Agentic Component Recognition (ACR)
Building Production-Ready Generative UI with Google’s Agent Development Kit
medium.com
June 30, 2025 at 6:20 PM
Adding the link helps:
medium.com/google-cloud...
medium.com/google-cloud...
Trying my hand at this YouTube thing, www.youtube.com/@stigsfoot
Noble Ackerson
There are over 100 million active YouTube channels. This channel is mine.
On here you may find content about #aibias, #evaluatingMLmodels, #privacy, #explainableai, #datagovernance, or #extendedrealit...
www.youtube.com
April 4, 2025 at 8:22 PM
Trying my hand at this YouTube thing, www.youtube.com/@stigsfoot
Model≠biz value By "real world" I am looking at e.g. token/compute cost<>value mapping.
U can't eliminate cost w/ local.
Also u still have the challenge of small models currently can't "reason" well, needs larger models. Also evals!!! The entropy of the output space is too high to be effective rn
U can't eliminate cost w/ local.
Also u still have the challenge of small models currently can't "reason" well, needs larger models. Also evals!!! The entropy of the output space is too high to be effective rn
February 21, 2025 at 7:16 PM
Model≠biz value By "real world" I am looking at e.g. token/compute cost<>value mapping.
U can't eliminate cost w/ local.
Also u still have the challenge of small models currently can't "reason" well, needs larger models. Also evals!!! The entropy of the output space is too high to be effective rn
U can't eliminate cost w/ local.
Also u still have the challenge of small models currently can't "reason" well, needs larger models. Also evals!!! The entropy of the output space is too high to be effective rn
Let me clarify. By hype is legit, I mean I can first hand confirm that these companies are hyping up a technology that's not ready for real world use.
February 20, 2025 at 2:50 AM
Let me clarify. By hype is legit, I mean I can first hand confirm that these companies are hyping up a technology that's not ready for real world use.
Right?
Then I see this in the news today...not surprising
Then I see this in the news today...not surprising
January 17, 2025 at 8:12 PM
Right?
Then I see this in the news today...not surprising
Then I see this in the news today...not surprising
A federal data privacy protection legislation, sir. That's the solution.
January 15, 2025 at 4:46 AM
A federal data privacy protection legislation, sir. That's the solution.