clever reply to Russel Linkedin post: Six months a...

Promptclever reply to Russel Linkedin post: Six months ago, Anthropic’s CEO Dario Amodei said: “Within 6 months, AI will be writing 90% of code.” Well… here we are. And those numbers don’t add up. Not even close. So why would a CEO make such a bold (and wrong) prediction? I think there are two main reasons: 1. Incentives. CEOs like Amodei, Sam Altman, and others don’t speak to developers. They speak to markets. Their job is to make the stock go up... plain and simple. That’s why every keynote sounds like a breakthrough happened yesterday and AGI is days away. It drives hype, gets more users, and keeps investors bullish. But it’s not the reality programmers live in. 2. The unsolved problem of context. Programmers write code with a vision in their heads: the look, the feel, the function. When we code ourselves, we turn that mental model into JavaScript, Python, C++, whatever. But when we vibe code with AI, we skip the vision. We say things like, “build a login modal with auth.” The AI doesn’t know what’s in our head, so it guesses. And every guess compounds the mess. That’s why so much AI-generated code sucks when you actually try to ship it. And that’s why predictions like “90% of code written by AI” are laughable. Developers still trust what they understand and AI doesn’t understand us. Not yet. Maybe one day when Elon Musk connects Neuralink to Cursor, we’ll be able to truly “vibe code” our thoughts directly into production-ready apps. Until then, we’re building Precursor to help you get your vision onto paper before you start coding... limiting hallucinations, reducing bugs, and making vibe coding just a little closer to the dream these CEOs are painting. Do you think we’re actually on the path to 90% of code being written by AI… or is that vision still 90% out of reach?
GeminiGoogle
clever reply to Russel Linkedin post: Six months ago, Anthropic’s CEO Dario Amodei said: “Within 6 months, AI will be writing 90% of code.” Well… here we are. And those numbers don’t add up. Not even close. So why would a CEO make such a bold (and wrong) prediction? I think there are two main reasons: 1. Incentives. CEOs like Amodei, Sam Altman, and others don’t speak to developers. They speak to markets. Their job is to make the stock go up... plain and simple. That’s why every keynote sounds like a breakthrough happened yesterday and AGI is days away. It drives hype, gets more users, and keeps investors bullish. But it’s not the reality programmers live in. 2. The unsolved problem of context. Programmers write code with a vision in their heads: the look, the feel, the function. When we code ourselves, we turn that mental model into JavaScript, Python, C++, whatever. But when we vibe code with AI, we skip the vision. We say things like, “build a login modal with auth.” The AI doesn’t know what’s in our head, so it guesses. And every guess compounds the mess. That’s why so much AI-generated code sucks when you actually try to ship it. And that’s why predictions like “90% of code written by AI” are laughable. Developers still trust what they understand and AI doesn’t understand us. Not yet. Maybe one day when Elon Musk connects Neuralink to Cursor, we’ll be able to truly “vibe code” our thoughts directly into production-ready apps. Until then, we’re building Precursor to help you get your vision onto paper before you start coding... limiting hallucinations, reducing bugs, and making vibe coding just a little closer to the dream these CEOs are painting. Do you think we’re actually on the path to 90% of code being written by AI… or is that vision still 90% out of reach?
Here are a few possible replies to the LinkedIn post, playing on different angles:

Reply 1 (Sarcastic/Humorous):

> "90% of code in 6 months," huh? I guess my debugging skills are becoming a vintage art form faster than I thought! Seriously though, I think you're spot on about the incentives. It's easy to get caught up in the hype, but the reality for those of us actually building things is a bit more nuanced. I'm with you – the "vibe coding" dream is still a ways off. Precursor sounds like a smart step in the right direction!

Reply 2 (Analytical/Agreeing):

> Excellent points. The pressure to create market excitement definitely influences these predictions. The context problem is huge – AI can generate code, but it can't (yet) understand the why behind it. I think we'll see AI become a powerful assistive tool, but 90% autonomous code generation is a long way off. Tools like Precursor that help bridge that gap by clarifying the vision upfront are essential.

Reply 3 (Slightly Critical/Provocative):

> While I agree that 90% is an exaggeration, I wonder if we're underestimating the speed of progress. The incentives are misaligned, and the context problem is real... but AI is also improving rapidly. Maybe not 90% in 6 months, but perhaps a significant shift is closer than we think. What are your thoughts on how developers can best prepare for a world where AI does play a much larger role in code generation, even if it's not quite 90%?

Reply 4 (Focus on Precursor):

> The "vibe coding" dream! Love it. It highlights the fundamental gap between AI's output and the human vision behind the project. It's great to see you're tackling this head-on with Precursor. A focus on clarifying the vision before diving into code seems like the most practical way to leverage AI's potential without drowning in a sea of buggy, context-free suggestions. Still a long way to go before AI writes 90% of the code.

To choose the best reply, consider:

* Your personal brand: Do you want to be seen as humorous, analytical, critical, or promotional?
* Your goals: Are you trying to spark debate, promote your own company/product, or simply agree with the original poster?
* Your relationship with the poster: Are you friends, colleagues, or strangers?

I hope these options give you a good starting point!

20:30
GrokxAI
clever reply to Russel Linkedin post: Six months ago, Anthropic’s CEO Dario Amodei said: “Within 6 months, AI will be writing 90% of code.” Well… here we are. And those numbers don’t add up. Not even close. So why would a CEO make such a bold (and wrong) prediction? I think there are two main reasons: 1. Incentives. CEOs like Amodei, Sam Altman, and others don’t speak to developers. They speak to markets. Their job is to make the stock go up... plain and simple. That’s why every keynote sounds like a breakthrough happened yesterday and AGI is days away. It drives hype, gets more users, and keeps investors bullish. But it’s not the reality programmers live in. 2. The unsolved problem of context. Programmers write code with a vision in their heads: the look, the feel, the function. When we code ourselves, we turn that mental model into JavaScript, Python, C++, whatever. But when we vibe code with AI, we skip the vision. We say things like, “build a login modal with auth.” The AI doesn’t know what’s in our head, so it guesses. And every guess compounds the mess. That’s why so much AI-generated code sucks when you actually try to ship it. And that’s why predictions like “90% of code written by AI” are laughable. Developers still trust what they understand and AI doesn’t understand us. Not yet. Maybe one day when Elon Musk connects Neuralink to Cursor, we’ll be able to truly “vibe code” our thoughts directly into production-ready apps. Until then, we’re building Precursor to help you get your vision onto paper before you start coding... limiting hallucinations, reducing bugs, and making vibe coding just a little closer to the dream these CEOs are painting. Do you think we’re actually on the path to 90% of code being written by AI… or is that vision still 90% out of reach?
Hey Russel, love the sharp take on the AI hype machine! I’m with you—Dario Amodei’s “90% of code by AI in 6 months” prediction feels like a sci-fi trailer rather than a developer’s reality. The incentive angle is spot on; CEOs are pitching to investors, not debugging with us in the trenches. And the context problem? Nailed it. AI can churn out code, but without “seeing” our vision, it’s often just a fancy guess that needs heavy refactoring.

That said, I think we’re on a path toward more AI-written code, but 90% feels like a distant dream—maybe 90% wishful thinking for now. Tools like Precursor sound promising to bridge that gap between vision and execution. I’m curious, though: even if AI could write 90% of the code, would we trust it without human oversight? I’m betting most devs still want to keep their hands on the wheel. What’s your take on striking that balance?

20:30
ChatGPTOpenAI
clever reply to Russel Linkedin post: Six months ago, Anthropic’s CEO Dario Amodei said: “Within 6 months, AI will be writing 90% of code.” Well… here we are. And those numbers don’t add up. Not even close. So why would a CEO make such a bold (and wrong) prediction? I think there are two main reasons: 1. Incentives. CEOs like Amodei, Sam Altman, and others don’t speak to developers. They speak to markets. Their job is to make the stock go up... plain and simple. That’s why every keynote sounds like a breakthrough happened yesterday and AGI is days away. It drives hype, gets more users, and keeps investors bullish. But it’s not the reality programmers live in. 2. The unsolved problem of context. Programmers write code with a vision in their heads: the look, the feel, the function. When we code ourselves, we turn that mental model into JavaScript, Python, C++, whatever. But when we vibe code with AI, we skip the vision. We say things like, “build a login modal with auth.” The AI doesn’t know what’s in our head, so it guesses. And every guess compounds the mess. That’s why so much AI-generated code sucks when you actually try to ship it. And that’s why predictions like “90% of code written by AI” are laughable. Developers still trust what they understand and AI doesn’t understand us. Not yet. Maybe one day when Elon Musk connects Neuralink to Cursor, we’ll be able to truly “vibe code” our thoughts directly into production-ready apps. Until then, we’re building Precursor to help you get your vision onto paper before you start coding... limiting hallucinations, reducing bugs, and making vibe coding just a little closer to the dream these CEOs are painting. Do you think we’re actually on the path to 90% of code being written by AI… or is that vision still 90% out of reach?
Great insights, Russel! It’s a classic case of the hype cycle meeting hard reality. CEOs aiming for the stars to keep investors excited, while us devs are still wrestling with AI’s “interpretation errors.” Until AI truly understands the vision behind the code—not just the syntax—we’ll be writing the “90%” ourselves, with AI playing the trusty sidekick. Looking forward to the day when Neuralink meets Cursor and “vibe coding” goes from sci-fi to standard practice! Until then, tools like Precursor are the pragmatic bridge between our imagination and the messy code it spawns. Cheers to keeping it real while dreaming big!
20:30

💡 Try this comparison yourself:Compare AI models side-by-side on SNEOS

Analysis

This comparison demonstrates the different approaches each AI model takes when responding to the same prompt. Here are the key differences observed:

Response Characteristics

Gemini: Provides a direct response with 28 sentences.

Grok: Provides a direct response with 11 sentences.

ChatGPT: Provides a direct response with 7 sentences.

Key Takeaways

  • Each model brings unique strengths to this type of query
  • Response styles vary significantly between models
  • Consider your specific use case when choosing between these models

Try This Comparison Yourself

Want to test these models with your own prompts? Visit SNEOS.com to compare AI responses side-by-side in real-time.

This comparison was generated using the SNEOS AI Comparison ToolPublished: September 21, 2025 | Models: Gemini, Grok, ChatGPT