Your Opinion Sucks...

What a great opportunity. Do both. Have them compete, spend the exact same money, run the same ads, and then REPORT your results. Having a test, a battle of the AI, with results would be a valuable report, and I mean this sincerely, MILLIONS of people would want to get their hands on. Do a 6 month trial, do exactly what each one says and let the market tell us the winner.

I'll be the first one in line to BUY your report. You are welcome for the many millions of dollars headed your way...cause with AI spewing such opposite advice, only a real life, real money head to head test with results would matter.

And the loser, the owners, would pay you MILLIONS not to publish your results.

GordonJ

I'm curious if the free version and the pro version would produce different results?

As an advocate of bringin' musself ahf au natchurelle when it comes to juicin' out the copy, I am no big fan of AI.

On the flip side, AI has advocates who regularly proclaim its glories.

So we gaht a slidin' scale here on an evocative lips ticket.

Pinpoint mersheen or smacko hooman bein'?

Gotta figure this is an unanswahered question for many -- which is why the kinda informed analysis Gordon presents as an offah sets evrywan straight on thullaytist dilemma.

Anyways, deal rn seems how huntaproneurial types can switch out the whole dream/tedium equation more productively.

"Quit the small losses to gain the big wins" kinda thing -- like mebbe franchisin' out yr BLINKIN' so's you can git on with SEEIN' an' PURSIEVIN'.

So here is Moi, mebbe 2240 ...

I clamped up to NipsyStim v23.9.

An' I gaht muh dreams set on bein' THUPPOPE.

So now AI servants somersault before my gaze.

Fixin' cash, networks, admin, outreach, intel -- plus HOW I still git to wear a knockout thong 'longside alla the historical hupparel.

See bcs AI gonna figure muh dream.

"Process My Toil While I Mostest My Inner Seem!"

Cray as cray.

An' so what an unveilityoore of possibility for evrywan.

Howevah, in my Sagittarian pernickityness, I would wish to pick up on the folly of the ubiquitous.

Naht jus' ovah an' ovah by default, but case by case.

"Should I plunge into a pool of acid?"

"You got skin problems, the fkr will surely sort them."

Which is why you gotta remembah 99%a nowan don't know nuthin'.

Evin' quotin' that sets you apart from the hapless -- speshy if'n you keep it to yusself.

Go back in a few days and you can get 2 different results from just one platform.

Just for the record, white background out performs colored backgrounds. An outlier in the commerce space would be AMAZON... but the text boxes are white, and I believe they change this back and forth from time to time ( the background shade and not the white text boxes )

Im going to sound like a grumpy old man, but open platform AI's are just short of trash ( in my opinion ) is this type of scenario - as well as many others.

I would gp a step further and check my analytics and also rin a site test to see what needs to be optimized and so on .

Use people for feed back too

well here's what i think . AI is basically a representation of a group of peoples training and thier perspective combned with general knowledge ( which all together is still peoples perspective ) . the two diffrent where trained diffrently and when it comes to creativity and design both have diffrent perspective .
this can be likened to peoples diffrent perspective on same things.

max5ty,

• Almost always. But not just different apps. The same models, used by different apps, too.
* And I regularly use ChatGPT and BeMyEyes for describing photos and other visual elements to me, among other general visual and non-visual querying tasks. These apps both use OpenAI models.

So where's the rub?
• These apps don't expose the system_prompt (for reasoning models, this is developer_prompt or similar). Just the user_prompt.
* Simply put, you aren't entering hardcoded directives. The app devs did that for you. And they aren't allowing you to change what they entered there. So with each dev, a difference in those hardcoded prompts is inevitable, given how unique and distinct each micro niche use case is, especially when considering target specific micro niche groups.

Why?
• I'd answer that first. Then That's part of finding the most suitable models for particular use cases.
* That just means data biases exist not just for your specific use case, but for almost all others, between what a dev believes is real about something as opposed to what your biases are, which significantly affects how the datasets used to train the model are developed and /or sourced by the dev, not you.

So the only way to have full deterministic control is to go opensource or build and train your own algorithms, using your own datasets.

I just started using AI myself. I also tried Gemini I like the speed of his answers but he has too much unnecessary info no like how he talks to me so chose to go back to ChatGPT still quite new in AI word maybe need check guides how use properly but ChatGPT sometimes so slow