Those tools are Godsent. Generally I only deal with hardware and let other people program it. AI has reduced my need to communicate with gammas, at least professionally, by as much as 90%. That's the real revolution.
Once again, great article. Please forgive me for asking a dumb question. If there is a hierarchy of coders and technologically savvy people, I’m at the bottom with the teeming masses, so these articles are incredibly helpful for sorting through the panic and unrealistic expectations and learning about the reality of AI and its potential.
Question: reading this, while these tools have been developed and can improve speed and efficiency for developers, they still require someone who knows what to look for to get a good product (code)? (Thinking of Vox Day and JDA using it to enhance their writing speed and stress testing mathematical probabilities.)
It’s the difference between knowing *how* to code versus knowing *what* to code. AI knows how to code so that won’t be a viable (human) job description anymore. For example, just knowing some programming language won’t be a job.
Consider the Antigravity approach: it has a human telling it what to code and the AI knows how to code it.
Practically speaking, I think that means programmers have to specialize. Become an expert in interface design, or information security, or whatever application so you can be the person that knows what the product should look like.
AI is being set up to replace the junior engineers and coders on a team. I don’t have a great answer as to how we get senior engineers when the junior engineers and the training pipeline are replaced by AI. One possibility is that “senior engineers” are born, not made. That is, it’s mostly a function of IQ, and AI replaces the bottom X% of people working in software.
Thank you. Would it potentially be similar to how we learn subjects like math and science? E.g. AI is in its infancy and we are stuck at its messy phase. Over time it will become more polished, mature, if you will. Eventually, it is seamlessly integrated into our daily lives, becoming something we manage if not fully understand? Similar to how we use the knowledge from previous generations of Greeks for mathematics and representative government, physics from 18th and 19th century scientists, or how today’s engineers don’t have to start from scratch when designing a car because they can refer back to the first inventors?
What you describe sounds like using AI as a library for knowledge. I think it’s more like AI will be used to replace people and jobs and labor. For example, if you want to write your will or some other legal document, you would have to hire a lawyer to do it. Instead, you will “hire” an AI to do it.
I have to try Cursor, thank you. Copilots biggest problem is it would create a debugging nightmare.
Those tools are Godsent. Generally I only deal with hardware and let other people program it. AI has reduced my need to communicate with gammas, at least professionally, by as much as 90%. That's the real revolution.
Once again, great article. Please forgive me for asking a dumb question. If there is a hierarchy of coders and technologically savvy people, I’m at the bottom with the teeming masses, so these articles are incredibly helpful for sorting through the panic and unrealistic expectations and learning about the reality of AI and its potential.
Question: reading this, while these tools have been developed and can improve speed and efficiency for developers, they still require someone who knows what to look for to get a good product (code)? (Thinking of Vox Day and JDA using it to enhance their writing speed and stress testing mathematical probabilities.)
Thank you.
It’s the difference between knowing *how* to code versus knowing *what* to code. AI knows how to code so that won’t be a viable (human) job description anymore. For example, just knowing some programming language won’t be a job.
Consider the Antigravity approach: it has a human telling it what to code and the AI knows how to code it.
Practically speaking, I think that means programmers have to specialize. Become an expert in interface design, or information security, or whatever application so you can be the person that knows what the product should look like.
AI is being set up to replace the junior engineers and coders on a team. I don’t have a great answer as to how we get senior engineers when the junior engineers and the training pipeline are replaced by AI. One possibility is that “senior engineers” are born, not made. That is, it’s mostly a function of IQ, and AI replaces the bottom X% of people working in software.
Thank you. Would it potentially be similar to how we learn subjects like math and science? E.g. AI is in its infancy and we are stuck at its messy phase. Over time it will become more polished, mature, if you will. Eventually, it is seamlessly integrated into our daily lives, becoming something we manage if not fully understand? Similar to how we use the knowledge from previous generations of Greeks for mathematics and representative government, physics from 18th and 19th century scientists, or how today’s engineers don’t have to start from scratch when designing a car because they can refer back to the first inventors?
What you describe sounds like using AI as a library for knowledge. I think it’s more like AI will be used to replace people and jobs and labor. For example, if you want to write your will or some other legal document, you would have to hire a lawyer to do it. Instead, you will “hire” an AI to do it.
Think of AI as a replacement for employees.
Yes. AI does not replace human talent; it supercharges those who learn to wield it.