Matt Bongiovi
@bongiovimatt.bsky.social
AI researcher focused on safety and explainability in language models
I have a webinar and I’m comfortable saying that, by attending, the very fabric of the universe will be re-woven.
Can you trust this grandiosity? Do you need supporting evidence for my dauntless hyperbole? Perhaps you want these claims to be backed up by…an explanation??
Well then sign on up.
#AI
Can you trust this grandiosity? Do you need supporting evidence for my dauntless hyperbole? Perhaps you want these claims to be backed up by…an explanation??
Well then sign on up.
#AI
Explainable AI in practice: techniques to build trust on top of LLMs
events.zoom.us
December 4, 2025 at 10:18 AM
I have a webinar and I’m comfortable saying that, by attending, the very fabric of the universe will be re-woven.
Can you trust this grandiosity? Do you need supporting evidence for my dauntless hyperbole? Perhaps you want these claims to be backed up by…an explanation??
Well then sign on up.
#AI
Can you trust this grandiosity? Do you need supporting evidence for my dauntless hyperbole? Perhaps you want these claims to be backed up by…an explanation??
Well then sign on up.
#AI
I work in AI, and I also don’t live in San Francisco. [the crowd gasps – shocked, but intrigued]
One of the nice things about this setup is that you get an outsider’s view. It’s possible to examine the cultural forces that influence decision-making.
Read more:
One of the nice things about this setup is that you get an outsider’s view. It’s possible to examine the cultural forces that influence decision-making.
Read more:
#ai #aisafety #chatgptformydog | Matt Bongiovi
I work in AI, and I also don’t live in San Francisco. [the crowd gasps – shocked, but intrigued]
One of the nice things about this setup is that you get an outsider’s view. It’s possible to examine t...
www.linkedin.com
December 1, 2025 at 10:22 AM
I work in AI, and I also don’t live in San Francisco. [the crowd gasps – shocked, but intrigued]
One of the nice things about this setup is that you get an outsider’s view. It’s possible to examine the cultural forces that influence decision-making.
Read more:
One of the nice things about this setup is that you get an outsider’s view. It’s possible to examine the cultural forces that influence decision-making.
Read more:
Justified Trust is a great concept, and one I think AI builders can utilize more. Pair engineers with compliance-savvy people so that you create the most socially-awkward conversations you’ve ever seen. Then, turn compliance obligations into meaningful design constraints.
#aisafety #ai | Matt Bongiovi
I find the term “Trust in AI” to be overused and vague and also bad for “reasons”. Saying "just trust us" doesn’t scale. Especially not in high-stakes situations like the ones UnlikelyAI operates in.
...
www.linkedin.com
November 19, 2025 at 11:14 AM
Justified Trust is a great concept, and one I think AI builders can utilize more. Pair engineers with compliance-savvy people so that you create the most socially-awkward conversations you’ve ever seen. Then, turn compliance obligations into meaningful design constraints.
If you’re like me and you have ears, you might have heard terms like Explainable AI, Trustworthy AI, Safe AI. These hint at a nicer future.
"But how do we get there", you whisper, praying your AI assistant can't hear you.
A compelling answer is “AI Assurance”.
bit.ly/4oxgGhm
"But how do we get there", you whisper, praying your AI assistant can't hear you.
A compelling answer is “AI Assurance”.
bit.ly/4oxgGhm
Introduction to AI assurance | Matthew Bongiovi
If you’re like me and you have ears, you might have heard a lot of terms like Explainable AI, Trustworthy AI, Safe AI, Fair AI …
These terms all get at the same idea: that there is a nicer future ah...
bit.ly
November 13, 2025 at 12:48 PM
If you’re like me and you have ears, you might have heard terms like Explainable AI, Trustworthy AI, Safe AI. These hint at a nicer future.
"But how do we get there", you whisper, praying your AI assistant can't hear you.
A compelling answer is “AI Assurance”.
bit.ly/4oxgGhm
"But how do we get there", you whisper, praying your AI assistant can't hear you.
A compelling answer is “AI Assurance”.
bit.ly/4oxgGhm