Uber's massive AI mistake revealed #tech #shorts
Uber's COO admitted the company cannot draw a clear line between heavy AI tool usage and increased useful customer features, sparking widespread 'AI bubble' narratives online. The speaker argues this interpretation misses the real story, contending that Uber is already doing meaningful agentic AI work and that the true bottleneck is compute power and token demand, not AI's usefulness.
Summary
The video addresses a widely circulating story about Uber's experience with AI coding tools. According to the speaker, Uber invested heavily in AI tools, saw engineers adopt them broadly, and measured rising token usage and AI-driven code commits. However, Uber's president and COO Andrew Macdonald publicly stated that the company still cannot establish a clear connection between all that AI activity and a measurable increase in useful customer-facing features.
This admission quickly became fuel for 'AI bubble' narratives across the internet, with commentators framing it as evidence that AI is too expensive, that token usage is backfiring, that AI agents aren't delivering value, and that the broader AI investment bubble is beginning to crack.
The speaker pushes back strongly against this interpretation, expressing frustration at what they see as a recurring misreading of AI's actual challenges. They argue the real bottleneck in AI development is not usefulness or ROI, but rather physical infrastructure — specifically power capacity and the ability to scale token throughput given how high demand already is. The speaker also disputes the framing that Uber has failed with AI agents specifically, claiming that public evidence actually shows Uber is already engaged in real, substantive agentic AI work, suggesting the popular narrative oversimplifies and distorts what the Uber story actually reveals.
Key Insights
- Uber's COO Andrew Macdonald stated that despite heavy AI tool usage, rising token counts, and increased AI-driven commits, the company still cannot draw a clean line to a clear increase in useful customer features.
- The speaker argues that the dominant online reaction — framing Uber's situation as evidence of an AI bubble cracking — is the wrong lesson to take from the story.
- The speaker contends that the real bottleneck in AI is not ROI or usefulness, but physical infrastructure: power capacity and the ability to serve more tokens given how high demand already is.
- The speaker claims that public evidence contradicts the narrative that Uber has failed with AI agents, asserting that Uber is already doing real agentic work.
- The speaker expresses repeated frustration that discussions about genuine AI infrastructure constraints get overshadowed by 'bubble' narratives whenever companies report unclear ROI from AI tool adoption.
Topics
Transcript
[0:00] Uber just gave everyone a very easy AI bubble headline. The company spent heavily on AI coding tools. Engineers used them widely. Token usage went up. AI-driven commits went up. And Uber's president and COO, Andrew Macdonald, said the company still cannot draw a clean line from all of that AI usage to a clear increase in useful customer features. That is the version of the story that's spreading everywhere online today. AI is too expensive. Token maxing is backfiring. Agents aren't worth the money. The bubble is cracking. I think [0:31] that's the wrong lesson, people. And I'm so tired of having to say that because I feel like the bubble people need to find more bubble stories…
Full transcript available for MurmurCast members
Sign Up to AccessMore from AI News & Strategy Daily | Nate B Jones
The AI skill nobody talks about (and it isn't prompting) #AI #prompting #productivity #tech
The key differentiator in AI productivity isn't prompting skills but the ability to write structured specifications that enable AI to function as an autonomous agent. A person with advanced specification skills can produce 10x more output than someone using basic prompting by investing upfront time in detailed requirements and then letting the AI work independently.
1.6M agents registered for OpenClaw and did NOTHING.
The speaker explains how to determine whether a task requires a single agent, multiple agents, a chat interface, or no AI at all by using four key estimation criteria. He addresses the failure of 1.6 million OpenClaw agents that were registered but unused, arguing the problem is matching tasks to appropriate solutions rather than a lack of tools.
The one question that tells you if your role is safe #AI #careers #AIjobs #jobs #tech
The speaker presents a critical question for evaluating job security in the age of AI: would your role still exist if the company were significantly smaller? If the answer is no, your value is tied to coordination rather than direct value creation, making your position vulnerable in leaner organizations. The solution is to migrate toward work that directly generates revenue and drives business direction while adopting engineering principles of precision, testability, and falsifiability.
When everyone can code, this is what's scarce #AI #careers #AIjobs #coding #tech
As AI coding capabilities become widespread, the critical skill shifts from writing code to translating business needs into precise specifications and validating whether solutions actually solve customer problems. The person who can bridge vague requirements and technical implementation while exercising judgment becomes the organization's center of gravity.
20 AI Agents Rebuilt My Wife's Website For $8. I Never Typed a Word.
A developer demonstrates how a multi-agent AI system rebuilt his wife's website in 1.5 hours for $8 by orchestrating cheaper models under a premium supervisor, catching four major failures (hallucinations, accessibility shortcuts, design bugs, and checker errors) without human intervention—achieving superior results compared to six days of single-agent work.