Only a short matter of time before agentic tools start serving ads too - paying user or not. You want to refactor your codebase ? No issue - taking 30 seconds - please view this ad meanwhile.
In my head, it'll be like the high pressure timeshare sales pitches or the dreaded car sales transactions, where they pull out all the tricks to convince you to buy something you don't actually want or need, regardless of whether you can afford it.
“That’s a great point about your finances. But did you know this company offers credit to someone in your position for only this low interest? I can apply on your behalf if you just sign this statement”
The code will come out just fine, so your atrophied brain will remain dependent on OliCorp's parasocial prosthesis for strenuous thinking, dissolving wariness in experiences of super productivity. Then elsewhere, when plausible, OliCorp will progressively nudge you in some direction sold as predefined weight bonus to third party customers. You won't even notice and really, isn't it a fair price for all that productivity? Of course, AI isn't always right yet, but I'd say in a very practical 95% of cases your goals and expectations are in alignment with OliCorp AI.
Don't forget, every website and service monetized automated access as a consequence of the AI scraping boom and made unauthorized web-indexing impossible, traditional search dried off. And when OliCorp finally turned off access to their legacy index monopoly in favor of AI interfacing, you really have no choice but to trust your friendly chat buddy. Who else are you gonna ask for the fix? Former friends, your family or neighbors? People you've grown to hate because sympathetic local clustering is discouraged through four color divisive information shaping. I mean, those guys really are at fault for your lack of self-efficacy, the hate is warranted. And you're too tired to bother anyway.
It’s worse than that. It will be more akin to ad placement in movies, except in this case it will slip this proprietary library into the suggested solution instead of that one. Or embed ads in the solution code.
Gemini kind of already does this. I use builder in AI studio and it tends to use Gemini in places where other solutions make more sense. Last week I needed users to input their address, validate it, and geo code it. Rather than using google maps APIs (way cheaper, free tier) it used Gemini to do it.
eh, maybe for the super vibey tools. I can't imagine anyone who wants to maintain the trust of devs would do this, they would so quickly pivot to something else, its not like the general public, where AI == chatgpt in their mind.
I highly doubt this one. These agents are pretty interchangeable, any one of them could decide to not show ads and steal huge market share. Programming is one of the few areas they can actually make gross margin, so ads would be a terrible business decision given the above. The ad revenue from it would be insignificant against the API calls / subscriptions.
Why stop there? You want to refactor your codebase? Sure, but you need to adopt this dependency and this cloud service.
Mixing Ads or sponsorships to influence LLMs is a really, really bad idea. Especially when they're competing with Search ... which means that for some, "AIs" are the only window into the world when looking something up.
> Sure, but you need to adopt this dependency and this cloud service.
THIS.
Asked to make an app using AWS? “I can do that, but have you considered the lower lifetime costs of using Azure? I can generate a configuration for AWS, Azure, or produce a price comparison table. Let me know which you prefer.”
Absolutely no business with any real money is going to be swayed to use one cloud provider over the other based on one vibe coded app.
There is so much other stuff that goes into why business make decisions about any large contract. I’m not in cloud sales. But I venture just close enough to the sun not to get burned
> Absolutely no business with any real money is going to be swayed
And what about everyone starting something? Or prototyping? And what if you don't have a choice: pay more or follow our sponsored guidelines?
This is a dangerous road without proper defences both in terms of legislation and policy (and I mean world-wide, world corps = world laws, not having to go to court in every country lol).
Also, end users need to be educated about all this because what is to stop John or Jane from uploading their receipts to GPT to make their taxes and ... oops "did you know you can switch insurance to XYZ" or ... AI browser proactively hiding content competing with their partners ... you looking for a healthcare package? The only one available is from our sponsor. Take it or leave it.
The tax situation already happens now with Intuit owning both Turbo Tax and CreditKarma to get you to sign up for credit cards.
If I were prototyping something and found I could do it cheaper somewhere else, I’m not sure I would be upset. I hate ads just for the bad user experience.
As long as it is clearly an ad and they say they have affiliations. It’s no different than what Google and Amazon does now.
But ironically enough, I was almost about to pay for Overcast years ago even though the author openly admitted that you didn’t get much of anything for it except for supporting him back then.
He then added a non slimy self hosted system to buy ads for other podcasts based on the category of podcast you were listening to at that very second (no tracking). I thought that was a great service.
I think I would actually lean into a tight integration between ChatGPT and something like booking.com[1], AirBNB, GetYourGuide, etc when looking for travel ideas.
[1] Well I personally wouldn’t because I am not as cost conscience as the average traveler and I value the loyalty programs and status of certain hotel chains and Delta airlines. But most travelers don’t and shouldn’t care.
But if they let me put in my loyalty numbers and book directly with Hyatt, Hilton and Delta, hell I might pay more for ChatGPT.
How many medium size migrations have you done? Its never that easy even if you are just hosting a bunch of VMs. Let alone if you are using any cloud specific services.
I have been involved in a few on the periphery working in cloud consulting (first at AWS itself and now an outside company). I actively avoid the “lift and shifts”. I come in for the “modernize” portion.
I have been either part of or opted out of well over a dozen - I have a policy of never leading “lift and shifts” (and never staff augmentation).
Once you actually sit down and come up with a project plan with your PMO, a cloud migration is hardly ever worth the effort unless the destination cloud provider is backing up a shit ton of money for not only operational credits but also for internal AWS Professional Services (where I worked when I was inside AWS) or an outside partner to help (current employer).
Hell even certain departments at Amazon would never go through the effort of migrating to AWS from the legacy CDO infrastructure.
The risk of regressions, the refactoring, the retraining, the politics, etc are hardly ever worth it.
> Once you actually sit down and come up with a project plan with your PMO [...] The risk of regressions, the refactoring, the retraining, the politics, etc are hardly ever worth it.
You are entirely correct, of course ... except that much of the management class simply does not care about any of those things.
Not to mention the sizable contingent of engineers will repeatedly get suckered by the pitch of "Just migrate all of your stuff to [shiny new thing] and all of your [reliable old thing] problems will go away" (a.k.a., "engineers with management potential")
Stripe already built this out and it’s in ChatGPT. Some merchants let you buy directly in the chat interface, and payment credentials are shared through the chat interface to the store (securely, etc)
