Human-in-the-loop
73
MENTIONS
25
EPISODES
24
PODCASTS
Search complete. 73 mentions across 25 episodes found for "Human-in-the-loop".
Sep 13, 2026
Ethics by Design: Building Trustworthy AI in Pharmaceutical Quality
S
28:22speaker_0HOST
There
S
28:22speaker_1HOST
must always be a human in a loop, or HITL.
S
28:25speaker_0HOST
HITL.
S
28:26speaker_1HOST
This means someone with the explicit authority and the technical interface to look at the AI's output, recognize a potential bias or error, and hit the brakes.
S
28:35speaker_0HOST
But hold on, if we build a super advanced AI that processes a million data points to make a recommendation and then we just force a human operator to look at it and say yes or no, aren't we just introducing a massive bottleneck?
The AI Scope Sentinel: Stop Unpaid Work Before It Starts
M
2:27Marcus ChenHOST
it never sends without your approval.
M
2:28Marcus ChenHOST
You're the human in the loop.
M
2:30Marcus ChenHOST
The AI just cuts down the time you spend analyzing every email.
M
2:33Marcus ChenHOST
You get a daily digest.
M
4:31Marcus ChenHOST
The AI isn't perfect.
M
4:32Marcus ChenHOST
It's a tool to surface patterns, not to make the final call.
M
4:36Marcus ChenHOST
That's why you're the human in the loop.
M
4:38Marcus ChenHOST
You review every alert and decide based on your relationship.
Ep 314: Stan Smith, Founder & CEO, Gradient AI
D
35:35David PrejeantHOST
That's the term that the data scientists tell me.
D
35:38David PrejeantHOST
Got to keep that human in the loop.
S
35:40Stan SmithGUEST
There's a couple ways to think about that.
S
35:41Stan SmithGUEST
So in the world of like small commercial, whether it's healthcare or commercial PNC, when the policy is very small and very limited amount of dollars, it's hard to justify having a human look at every single one.
"Human in the Loop" Might Be the Worst Way to Use AI, with Vivienne Ming
V
11:27Vivienne MingGUEST
Their brain goes from solving hard math problem to chilling out in front of the TV.
V
11:32Vivienne MingGUEST
Then we had another group, let's call them human in the loop.
V
11:35Vivienne MingGUEST
You hear this a lot.
V
11:36Vivienne MingGUEST
Human in the loop often was worse than human all by themselves.
V
11:40Vivienne MingGUEST
Because human in the loop, when we analyze our students doing this, and I guarantee you, this is also true in the workplace, it means you assign an AI to do part of the task.
V
11:52Vivienne MingGUEST
Like, go analyze historical oil prices.
V
11:57Vivienne MingGUEST
And then it comes back and it gives you an answer.
V
12:05Vivienne MingGUEST
You either accept it, in which case you're just like my first group, Or you reject it, in which case you're like a human without an AI.
AI, Compliance & Thinking Outside the Box
H
23:31Hema LomaxHOST
And Audrey, may I just say, psychological safety is a problem if what you want to be safe to be is the same status quo, right? The protecting you and your fear is not the safety we're talking about.
H
23:40Hema LomaxHOST
I think what we need to be is safe to learn a growth mindset and get that human in the loop to do that.
H
23:44Hema LomaxHOST
So Mara, over to you.
H
23:46Hema LomaxHOST
Tell us about how we can use a human in the loop.
H
23:48Hema LomaxHOST
I know that's the foundation of Ethico, so one of them.
H
23:50Hema LomaxHOST
Tell us how that works for you.
AI Is Coming for Surgical Training — Here's What Happens Next with Samir Grover
S
50:10Samir GroverGUEST
Attach it onto its own email, make it read your emails.
S
50:15Samir GroverGUEST
And if there's simple things that need to be done that are HITL through a human in the loop through you in terms of deciding what those are, you could trigger things.
S
50:24Samir GroverGUEST
It can organize files brilliantly.
S
50:26Samir GroverGUEST
It can organize mailboxes brilliantly.
AI Should Give You More Capacity to Be Human
R
8:18Rachel Shields-WilliamsGUEST
Um, one of the groups just intuitively took a more sophisticated problem, and instead of trying to tackle it in one fell swoop, they actually broke it into, um, four or five separate agents so they could make the judgment call in between each agent.
R
8:35Rachel Shields-WilliamsGUEST
And they're like, "I mean, we might have made it more complicated." I was like, "No, actually, intuitively you made the right decision." I'm like, "You created the opportunity for the human to intervene and say, 'Is this the right step?'" And that is what we, that is human in the loop.
R
8:48Rachel Shields-WilliamsGUEST
Um, you know, you have the agent take a step.
R
8:51Rachel Shields-WilliamsGUEST
You look at it.
Episode 139 - Focusing on staffing margins could be costing you millions
D
38:05David MasonGUEST
And I think it's also this stage, you know, it's careful design.
D
38:13David MasonGUEST
None of this technology and the technologists and those involved in organizational design will talk about human in the loop or human on the loop.
D
38:24David MasonGUEST
But imagine you've got a two-year-old and you've just trained it to throw a ball.
D
38:28David MasonGUEST
It's still going to need practice and coaching to throw it better.
Everyone Has an LLM. Not Everyone Has the Data with Angie Judge
A
0:34Angie JudgeGUEST
does it mean to have a human in the
H
0:36Hamish FriedlanderHOST
loop? Human in the loop.
H
0:36Hamish FriedlanderHOST
Human in the loop.
H
0:37Hamish FriedlanderHOST
What in the world is a human in the loop? Human in the loop.
H
0:41Hamish FriedlanderHOST
AI in the loop.
H
0:42Hamish FriedlanderHOST
Humans in
A
0:43Angie JudgeGUEST
charge.
A
0:44Angie JudgeGUEST
Welcome to Human in the Loop.
Regulating AI-Created Devices
R
10:21Ravi NabarGUEST
It's also, there's an acronym for it.
R
10:23Ravi NabarGUEST
It's called HITL.
R
10:24Ravi NabarGUEST
It's human in the loop.
R
10:26Ravi NabarGUEST
And so it's important to know that the human being can actually observe the things that AI wants it to observe or that we want observed.
15 more episodes mention Human-in-the-loop.
Create an account to see the whole feed, search across every transcript, and follow the entities you care about.