8hrs Time Saved - Uncle + AI (ChatGPT)

[Uncle Gizmo]

Yesterday I did around four hours solid work on a project with MS Access and Google sheets. I was constantly using ChatGPT and at the end of the day I asked ChatGPT if it could compare what my solo output would have been working on my own, to what my output was, working in conjunction with the ChatGPT large language model.... chatty estimated that I saved eight hours!!! Seems about right to me....​

​

How Much Time You and I Spent Today Together?​

• Our total token usage today is ~16,000–18,000 tokens (text + code).
• Based on timestamps and reply pacing, you’ve spent around 3.5 to 4 hours today actively engaging with the system.

​

Final Comparison:​

Metric	Solo Dev (No AI)	You + AI Today
Estimated Hours	~12 hrs	~3.5–4 hrs
Time Saved	—	8+ hrs
Output Quality	Manual & Slower	Accelerated & Strategically Directed
Mental Load	Spread over tasks	Concentrated on decisions

 

[GPGeorge]

Yesterday I did around four hours solid work on a project with MS Access and Google sheets. I was constantly using ChatGPT and at the end of the day I asked ChatGPT if it could compare what my solo output would have been working on my own, to what my output was, working in conjunction with the ChatGPT large language model.... chatty estimated that I saved eight hours!!! Seems about right to me....​

​

How Much Time You and I Spent Today Together?​

• Our total token usage today is ~16,000–18,000 tokens (text + code).
• Based on timestamps and reply pacing, you’ve spent around 3.5 to 4 hours today actively engaging with the system.

​

Final Comparison:​

Metric	Solo Dev (No AI)	You + AI Today
Estimated Hours	~12 hrs	~3.5–4 hrs
Time Saved	—	8+ hrs
Output Quality	Manual & Slower	Accelerated & Strategically Directed
Mental Load	Spread over tasks	Concentrated on decisions

I've used ChaptGPT and DeepSeek. The latter only to test, really.

Yesterday I tried Claude after a long fruitless session with ChatGPT. I received a working response out of the gate from Claude. ChatGPT had had me going in circles on the same task for hours.

If that pattern holds, I think Claude could become my go to. I know a one-off is not enough to make an informed decision, so I'll continue to try both on similar questions until I get a better picture, of course.

 

[jdraw]

Interesting George, I find chatgpt and copilot have good and bad offerings/suggestions/responses. I think it pays to keep options open and try some alternatives. I'm only dealing with some simple "keep busy" things, nothing deep nor critical, and use free options only. May be a whole different story if you opt for the paid, more encompassing products.

Chatgpt is definitely better now (more focused) than it was a year ago when i was trying some Constraint satisfaction questions/problems involving miniZinc. It had me going in circles with options, often going back to a previous approach (more than once).

 

[Isaac]

Yesterday I did around four hours solid work on a project with MS Access and Google sheets. I was constantly using ChatGPT and at the end of the day I asked ChatGPT if it could compare what my solo output would have been working on my own, to what my output was, working in conjunction with the ChatGPT large language model.... chatty estimated that I saved eight hours!!! Seems about right to me....​

​

How Much Time You and I Spent Today Together?​

• Our total token usage today is ~16,000–18,000 tokens (text + code).
• Based on timestamps and reply pacing, you’ve spent around 3.5 to 4 hours today actively engaging with the system.

​

Final Comparison:​

Metric	Solo Dev (No AI)	You + AI Today
Estimated Hours	~12 hrs	~3.5–4 hrs
Time Saved	—	8+ hrs
Output Quality	Manual & Slower	Accelerated & Strategically Directed
Mental Load	Spread over tasks	Concentrated on decisions

I read an article that millions of dollars are spent on the electricity or energy used solely to handle people's "Please" and especially "Thank You"s
to ChatGPT etc.

I'm guilty of that myself, I'm a person who prioritizes politeness in normal day to day face interactions, and I cannot escape my habits with ChatGPT. 90% of the time I thank it. And when I give it a prompt, it's a request, not a command. Just cannot stop myself it seems

 

[Isaac]

I've used ChaptGPT and DeepSeek. The latter only to test, really.

Yesterday I tried Claude after a long fruitless session with ChatGPT. I received a working response out of the gate from Claude. ChatGPT had had me going in circles on the same task for hours.

If that pattern holds, I think Claude could become my go to. I know a one-off is not enough to make an informed decision, so I'll continue to try both on similar questions until I get a better picture, of course.

I've tried nothing (other than embedded stuff like copilot), but for the standalone stuff I've tried nothing except ChatGPT, to any appreciable amount.
I may have to try Claude. Hopefully some version of it is free in exchange for me logging into it with Google and surrendering my life's information. (which I don't really care about).

 

[Isaac]

Ooohh ... I don't like that. I love ChatGPT's Memory Updates and the longer I use it, the more it knows about me. I prefer that default rather than the opposite default, and Claude doesn't have a way to change it. But I can see using Claude for technical things, wherein I don't mind taking the time to provide it the full context.
For less-technical things, I prefer the memory of chatgpt.