StarZero All shows Try the podcast clipper →

Apr 7, 2025 · 28m · big-technology

Understanding Practical AI and the Future of Automation – With Joseph George

Joseph George · 17m spoken Alex Kantrowitz · 8m spoken
0:00 / 0:00
▶ Watch on YouTube →

gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions

In this interview, Alex Kantrowitz speaks with GoTo's Joseph George about how generative AI and virtual technicians are transforming IT support from reactive manual troubleshooting into proactive, automated workflows. They explore practical enterprise implementations, the vital role of human-in-the-loop oversight, and the evolving technological architecture powering modern IT management.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Alex holds 30.7% of the talking time here. How this is scored →

Alex as informed peer 4.7 Guest teaching 2.5 Guest disagreement 0.2 Alex pushing back 1.2
05100:0010:0020:001:49–6:40 · Alex as informed peer 4/10 Automating Session Summaries and Deploying Virtual Technicians Alex effectively synthesizes the difference between doing IT work and documenting it. George explains how LogMeIn uses generative AI to convert session logs into executable automation scripts for virtual technicians.6:40–12:37 · Alex as informed peer 5/10 Human-in-the-Loop Supervision and Automated Workflow Execution Alex draws parallels to medical documentation and probes whether the system operates autonomously. George details how LLMs build resilient execution workflows under human supervision.12:37–17:02 · Alex as informed peer 4/10 Transitioning from Reactive Troubleshooting to Proactive IT Management Alex jokingly asks if the tech actually works and references conversations with industry leaders like Zendesk's CEO. George outlines the transition from reactive ticketing to proactive, telemetry-driven automated fixes.17:04–21:47 · Alex as informed peer 5/10 Workforce Evolution, Junior IT Roles, and Human-AI Collaboration Alex challenges the optimistic workforce framing by highlighting the risk to entry-level IT roles and citing developer over-reliance on tools like Cursor. George responds using Erik Brynjolfsson's framing of human-AI collaboration.21:47–25:16 · Alex as informed peer 5/10 Architecture, Foundation Models, and Product Design Considerations Alex drills into the tech stack and references specific frontier models while asking whether humans must train every single path. George shakes his head and clarifies that models generalize and generate scripts autonomously.25:18–27:23 · Alex as informed peer 5/10 Assessing the AI Plateau Debate and Emerging Small Language Models Alex brings up the debate around pre-training hitting a wall and whether AI is in a bubble. George counters that small language models and reasoning architectures mean enterprise AI is still in its earliest stages.1:49–6:40 · Guest teaching 3/10 Automating Session Summaries and Deploying Virtual Technicians Alex effectively synthesizes the difference between doing IT work and documenting it. George explains how LogMeIn uses generative AI to convert session logs into executable automation scripts for virtual technicians.6:40–12:37 · Guest teaching 3/10 Human-in-the-Loop Supervision and Automated Workflow Execution Alex draws parallels to medical documentation and probes whether the system operates autonomously. George details how LLMs build resilient execution workflows under human supervision.12:37–17:02 · Guest teaching 3/10 Transitioning from Reactive Troubleshooting to Proactive IT Management Alex jokingly asks if the tech actually works and references conversations with industry leaders like Zendesk's CEO. George outlines the transition from reactive ticketing to proactive, telemetry-driven automated fixes.17:04–21:47 · Guest teaching 2/10 Workforce Evolution, Junior IT Roles, and Human-AI Collaboration Alex challenges the optimistic workforce framing by highlighting the risk to entry-level IT roles and citing developer over-reliance on tools like Cursor. George responds using Erik Brynjolfsson's framing of human-AI collaboration.21:47–25:16 · Guest teaching 2/10 Architecture, Foundation Models, and Product Design Considerations Alex drills into the tech stack and references specific frontier models while asking whether humans must train every single path. George shakes his head and clarifies that models generalize and generate scripts autonomously.25:18–27:23 · Guest teaching 2/10 Assessing the AI Plateau Debate and Emerging Small Language Models Alex brings up the debate around pre-training hitting a wall and whether AI is in a bubble. George counters that small language models and reasoning architectures mean enterprise AI is still in its earliest stages.1:49–6:40 · Guest disagreement 0/10 Automating Session Summaries and Deploying Virtual Technicians Alex effectively synthesizes the difference between doing IT work and documenting it. George explains how LogMeIn uses generative AI to convert session logs into executable automation scripts for virtual technicians.6:40–12:37 · Guest disagreement 0/10 Human-in-the-Loop Supervision and Automated Workflow Execution Alex draws parallels to medical documentation and probes whether the system operates autonomously. George details how LLMs build resilient execution workflows under human supervision.12:37–17:02 · Guest disagreement 0/10 Transitioning from Reactive Troubleshooting to Proactive IT Management Alex jokingly asks if the tech actually works and references conversations with industry leaders like Zendesk's CEO. George outlines the transition from reactive ticketing to proactive, telemetry-driven automated fixes.17:04–21:47 · Guest disagreement 0/10 Workforce Evolution, Junior IT Roles, and Human-AI Collaboration Alex challenges the optimistic workforce framing by highlighting the risk to entry-level IT roles and citing developer over-reliance on tools like Cursor. George responds using Erik Brynjolfsson's framing of human-AI collaboration.21:47–25:16 · Guest disagreement 1/10 Architecture, Foundation Models, and Product Design Considerations Alex drills into the tech stack and references specific frontier models while asking whether humans must train every single path. George shakes his head and clarifies that models generalize and generate scripts autonomously.25:18–27:23 · Guest disagreement 0/10 Assessing the AI Plateau Debate and Emerging Small Language Models Alex brings up the debate around pre-training hitting a wall and whether AI is in a bubble. George counters that small language models and reasoning architectures mean enterprise AI is still in its earliest stages.1:49–6:40 · Alex pushing back 0/10 Automating Session Summaries and Deploying Virtual Technicians Alex effectively synthesizes the difference between doing IT work and documenting it. George explains how LogMeIn uses generative AI to convert session logs into executable automation scripts for virtual technicians.6:40–12:37 · Alex pushing back 1/10 Human-in-the-Loop Supervision and Automated Workflow Execution Alex draws parallels to medical documentation and probes whether the system operates autonomously. George details how LLMs build resilient execution workflows under human supervision.12:37–17:02 · Alex pushing back 2/10 Transitioning from Reactive Troubleshooting to Proactive IT Management Alex jokingly asks if the tech actually works and references conversations with industry leaders like Zendesk's CEO. George outlines the transition from reactive ticketing to proactive, telemetry-driven automated fixes.17:04–21:47 · Alex pushing back 2/10 Workforce Evolution, Junior IT Roles, and Human-AI Collaboration Alex challenges the optimistic workforce framing by highlighting the risk to entry-level IT roles and citing developer over-reliance on tools like Cursor. George responds using Erik Brynjolfsson's framing of human-AI collaboration.21:47–25:16 · Alex pushing back 1/10 Architecture, Foundation Models, and Product Design Considerations Alex drills into the tech stack and references specific frontier models while asking whether humans must train every single path. George shakes his head and clarifies that models generalize and generate scripts autonomously.25:18–27:23 · Alex pushing back 1/10 Assessing the AI Plateau Debate and Emerging Small Language Models Alex brings up the debate around pre-training hitting a wall and whether AI is in a bubble. George counters that small language models and reasoning architectures mean enterprise AI is still in its earliest stages.

speaking balance: gold is Alex, purple is the guest (3 minute bins)

0:00 · Alex 35.6% · guest 64.4%0:00 · Alex 35.6% · guest 64.4%3:00 · Alex 32.5% · guest 67.5%3:00 · Alex 32.5% · guest 67.5%6:00 · Alex 41.9% · guest 58.1%6:00 · Alex 41.9% · guest 58.1%9:00 · Alex 25.5% · guest 74.5%9:00 · Alex 25.5% · guest 74.5%12:00 · Alex 17% · guest 83%12:00 · Alex 17% · guest 83%15:00 · Alex 45.5% · guest 54.5%15:00 · Alex 45.5% · guest 54.5%18:00 · Alex 22.7% · guest 77.3%18:00 · Alex 22.7% · guest 77.3%21:00 · Alex 18% · guest 82%21:00 · Alex 18% · guest 82%24:00 · Alex 31.5% · guest 68.5%24:00 · Alex 31.5% · guest 68.5%27:00 · Alex 43.7% · guest 56.3%27:00 · Alex 43.7% · guest 56.3%
Sharpest disagreement ▶ 23:12 Premise Clarification on Model Training

In a very collegial conversation, George shakes his head to dismiss the idea that models need human walkthroughs for every scenario.

 Hardest push from Alex ▶ 17:04 Challenging Workforce Optimism

Alex refuses to accept corporate platitudes and presses George on whether entry-level IT workers will lose the ability to learn on the job.

 Biggest teaching moment ▶ 4:53 Session Summaries to Parallel Execution

George educates Alex on how modern IT AI moves beyond simple documentation to automatically generate and execute script workflows across multiple machines.

 Alex holds their own ▶ 23:53 Product vs Model Frontier Debate

Alex demonstrates industry fluency by framing the model versus product debate and citing specific model transitions like Claude Sonnet 3.7.
the scores for every segment, with the reasoning behind each
ChapterTopicAlex as informed peerGuest teachingGuest disagreementAlex pushing backWhy
Automating Session Summaries and Deploying Virtual Technicians 4300 Alex effectively synthesizes the difference between doing IT work and documenting it. George explains how LogMeIn uses generative AI to convert session logs into executable automation scripts for virtual technicians.
Human-in-the-Loop Supervision and Automated Workflow Execution 5301 Alex draws parallels to medical documentation and probes whether the system operates autonomously. George details how LLMs build resilient execution workflows under human supervision.
Transitioning from Reactive Troubleshooting to Proactive IT Management 4302 Alex jokingly asks if the tech actually works and references conversations with industry leaders like Zendesk's CEO. George outlines the transition from reactive ticketing to proactive, telemetry-driven automated fixes.
Workforce Evolution, Junior IT Roles, and Human-AI Collaboration 5202 Alex challenges the optimistic workforce framing by highlighting the risk to entry-level IT roles and citing developer over-reliance on tools like Cursor. George responds using Erik Brynjolfsson's framing of human-AI collaboration.
Architecture, Foundation Models, and Product Design Considerations 5211 Alex drills into the tech stack and references specific frontier models while asking whether humans must train every single path. George shakes his head and clarifies that models generalize and generate scripts autonomously.
Assessing the AI Plateau Debate and Emerging Small Language Models 5201 Alex brings up the debate around pre-training hitting a wall and whether AI is in a bubble. George counters that small language models and reasoning architectures mean enterprise AI is still in its earliest stages.

Statements from this episode (7)

Assertion Supported
GoTo: LogMeIn Resolve deploys parallel virtual technicians supervised by humans today
“And so the world of having a human technician now actually supervising what I call virtual technicians that can execute commands on their behalf and run in parallel on a number of different machines at the same time. That world sounds a little bit like science…”
Joseph George Apr 7, 2025 ▶ 5:46
Assertion Supported
George: LLMs Dynamically Generate and Adapt IT Automation Scripts
“Instead, the LLM is automatically generating that execution plan so it can be run. And even better, as you run that script the LLM is also determining if you encounter Behavior that you didn't expect. It's resilient enough to be able to work around that.”
Joseph George Apr 7, 2025 ▶ 9:02
Prediction Not checkable as stated
George: Virtual IT Agents Will Gain Autonomy as Success Rates Are Proven
“I can see the virtual agents taking on more autonomy over time, especially if these are proven types of problems and you understand and you've got a success rate of solving them. Then you could reach a case where there's even more automation happening there as…”
Joseph George Apr 7, 2025 ▶ 10:37
Prediction Not checkable as stated
George: IT systems will fix issues before users report them
“We're not there yet where everything's happening automatically, but we've got all the pieces and we're putting those together and that's the world we're getting to firmly where You don't even have to call in and talk about the problem. The system detects that …”
Joseph George Apr 7, 2025 ▶ 13:57
Prediction Not checkable as stated
George: Maturing AI agents will perform specialized rather than generic tasks
“The role of AI agents as that space matures, you'll see them doing more specific tasks rather than generic ones.”
Joseph George Apr 7, 2025 ▶ 21:04
Prediction Not checkable as stated
George: Practical federated learning for IT AI remains a long way off
“Over time as well, I think we'll get to a world where federated learning is also an option. But obviously as you think about federated learning, where you're learning from one set of experiences and transcribing that to another environment you've got to also m…”
Joseph George Apr 7, 2025 ▶ 22:46
Insight
George: Product design and human-model handoffs matter as much as AI models
“The product is just as important as well, right? Because how the user interacts with the system where models are leveraged and you have the right sort of handoff from human being to the model where you've got the ability, you know, think of it from a user inte…”
Joseph George Apr 7, 2025 ▶ 24:45
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 300 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.

Try the podcast clipper Learn more