Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Predicting OpenAI's ad strategy

ossa-ma.github.io

Post reply on HN
Print view View original

371–380 of 552 posts

Re: Predicting OpenAI's ad strategy

#371

Post by Forgeties79 »

littlecranky67 wrote: post #322
asdff wrote: post #321

Earlier quoted context omitted.

Sad truth is vast majority of users today are not savvy enough to install a good adblocker. We live in a bubble here in technology where we assume people know to do things like block ads. Really at this point people don’t know what files are anymore or even what a website actually is.

True, but some 20 years ago, I already could (and did) use photoshop to cut out the head of a woman (celebrity or from my social group) and put it onto a naked or bikini body of a woman. This is an old problem, but the bar to achieve that was too high and required tech savyness. The second Grok does it on twitter available for every idiot, suddently politicians see this as problematic and want to pass laws against it…

The problem is the way consumer rights are going, companies will simply make it impossible for you to run a line like that, or at least make it difficult enough that only a small percentage of the population can do it.

Every layer of friction they add means millions of people just throwing up their hands and accepting whatever nonsense is thrown at them.

Re: Predicting OpenAI's ad strategy

#372

Post by djhn »

Aurornis wrote: post #211
_factor wrote: post #200

Earlier quoted context omitted.

I doubt the ROI would be so high if organic results stood any chance. The ROI on bribes is very high too, but we haven’t legalized those (officially).

> I doubt the ROI would be so high if organic results stood any chance. This is just the same fallacy. In what world are people going to organically share ads for this company on their Facebook feeds? Who is going to Google the company name before they know about it? Every business needs to proactively acquire customers. Distribution and CAC are top of mind values for any growth business. It has been this way long be…

CAC-LTV by itself is not a great long term strategy. But it can appear like one in the short term to investors.

So there is a plethora of companies happily dumping investor money into paid customer acquisition.

Blue Apron was a classic example of borderline fraudulent growth metrics and a cash grab IPO before the customer acquisition funnel collapsed.

Re: Predicting OpenAI's ad strategy

#373

Post by littlecranky67 »

Forgeties79 wrote: post #371
littlecranky67 wrote: post #322

Earlier quoted context omitted.

True, but some 20 years ago, I already could (and did) use photoshop to cut out the head of a woman (celebrity or from my social group) and put it onto a naked or bikini body of a woman. This is an old problem, but the bar to achieve that was too high and required tech savyness. The second Grok does it on twitter available for every idiot, suddently politicians see this as problematic and want to pass laws against it…

The problem is the way consumer rights are going, companies will simply make it impossible for you to run a line like that, or at least make it difficult enough that only a small percentage of the population can do it. Every layer of friction they add means millions of people just throwing up their hands and accepting whatever nonsense is thrown at them.

You assume all models will be closed-source and there will be no competition. I can imagine that, but I would assume training an AI model for ad-filtering is something the community can and will do - maybe think 5 years ahead what hardware will be available for training and inference to a "small developer".

Re: Predicting OpenAI's ad strategy

#374

Post by mx20 »

Quarrelsome wrote: post #82
barbazoo wrote: post #68

Earlier quoted context omitted.

I’m assuming they have much more control during training and at runtime than us with our prompts. They’ll bake in whatever the person with the checkbook says to.

if they want dynamic pricing like adwords then its going to be a little challenging. While I appreciate its probably viable and they employ very clever people there's nothing like doing two things that are basically diametrically opposed at the same time. The LLM wants to give you what _should_ be the answer, but the winner of the ad word wants something else. There's a conflict there that I'd imagine might be quite…

Generate an Answer, get the winning Ad from an API, let another AI rewrite the Answer in a easy that the Answer at least be not contradicting to the Ad.

I think someone should create a leaderboard that measures how much the AI is lying to us to sell more ads.

Re: Predicting OpenAI's ad strategy

#375

Post by dang »

emil-lp wrote: post #362
dang wrote: post #307

Earlier quoted context omitted.

This is standard HN moderation. The title was linkbait, so we changed it. From https://news.ycombinator.com/newsguidelines.html : " Please use the original title, unless it is misleading or linkbait ". Note the unless. When we do this, we try to find a representative phrase from the article itself that accurately and neutrally represents what the article is about. There is nearly always one of those if you look for i…

Thank you for your service, dang. You do a great job keeping HN a great site. I disagree with your decision regarding changing the title, though. Yes, it's tongue-in-cheek, but I wouldn't call it bait. I wonder how HN anno 1968 would have moderated "Go To Statement Considered Harmful". * * Amusing side note: Dijkstra's submitted title was actually "A Case Against the Goto Statement" but the editor editorialized it.

As I read or rather hear it, "The A in AGI stands for Ads" has two things qualifying it as linkbait in the HN sense. One is snark, the other is the mammoth meme ("AGI"). Both those things stick to the Big Ball of Mud that rolls chaotically over most of the internet.

The objection here isn't to wit, it's to predictability. I'm not sure what the equivalent of internet snark was in 1968, but I doubt that "Go To Statement Considered Harmful" embodied it.

Of course, interpretations differ. No two people would agree on all these classifications, including me and tomhow. The important point isn't that we always make the correct call (we don't, plus there's no such thing as the correct call), but that we try to the principles from the guidelines, even though doing so involves interpretation and is therefore uneven.

On the bright side, it means that objections to titles and/or title edits, always good for firing up an HN thread, will always be with us!

https://news.ycombinator.com/item?id=8770124 (Dec 2014)

Re: Predicting OpenAI's ad strategy

#376

Post by Forgeties79 »

jeron wrote: post #191
yoyohello13 wrote: post #187

The ad based business model is the most destructive technology of the new millennium. The whole point of advertising, to make people feel unsatisfied with their current life. The current state of the world makes perfect sense when you think of the last 20 years we've been cultivating fear, and dissatisfaction among the populace. All in the hopes of selling 0.1% more widgets.

you thought the business model is any different in the 20th century? I recommend you watch the TV series Mad Men

The tools and information available back in the 60’s are not remotely comparable to the ones available today.

Re: Predicting OpenAI's ad strategy

#377

Post by Forgeties79 »

littlecranky67 wrote: post #373
Forgeties79 wrote: post #371

Earlier quoted context omitted.

The problem is the way consumer rights are going, companies will simply make it impossible for you to run a line like that, or at least make it difficult enough that only a small percentage of the population can do it. Every layer of friction they add means millions of people just throwing up their hands and accepting whatever nonsense is thrown at them.

You assume all models will be closed-source and there will be no competition. I can imagine that, but I would assume training an AI model for ad-filtering is something the community can and will do - maybe think 5 years ahead what hardware will be available for training and inference to a "small developer".

I don’t assume that at all. I’m saying that they just need to capture the vast majority of people, which has happened many times. For instance, Linux is available to everyone and would end a lot of frustrations and headaches they currently experience with macOS and windows. But for many reasons it still remains in the low single digit percentages when it comes to adoption, a lot of which is inflated by the steam deck.

It’s the same reason people pay to get their oil changed rather than doing it themselves. We are willing to compromise for convenience. And for a lot of people, that means just not dealing with it. Accept the ads, pay a small fee, whatever is the quickest way to make the thing works again or to get where they need to go.

Re: Predicting OpenAI's ad strategy

#378

Post by ElFitz »

prmoustache wrote: post #167
fsflover wrote: post #72

Earlier quoted context omitted.

Perhaps the actual solution is to ban ads through regulation, https://news.ycombinator.com/item?id=43595269

I am all for it but it would be hard to enforce. Ads are already hidden everywhere. Someone "review" a product? Most of the time it is a hidden ad even if the reviewer hasn't been paid with money for that. Watches a movie or a video clip? A lots of products are advertised through close ups on the logos, etc.

Even a sign on the street is basically an ad. That plastic cow or pig head telling you there’s a butcher or cheesemonger?

It’s basically an ad. Possibly one of the oldest type too!

Re: Predicting OpenAI's ad strategy

#379

Post by mupuff1234 »

bearjaws wrote: post #62

There is something to be said about the state of advertising. Increasingly it seems you must go to the almighty Google or Meta in order to launch any business. We're looking to expand into a new business line and have out grown our pharmacy capacity. The new business line will cost about $2M in software dev, and $3M for the new facility. The advertising budget? $40,000,000 (annual). We can build 10 robotic pharmacies…

Is your product offering anything unique to the consumer other than blasting them with more ads than the competition?

Re: Predicting OpenAI's ad strategy

#380

Post by nindalf »

johnnienaked wrote: post #366
Aurornis wrote: post #351

Earlier quoted context omitted.

What big picture? Can you elaborate, or just snide drive-by criticisms? I mean I understand that a lot of people in this comment section wish ads didn’t exist and that everyone just automatically knew about relevant products and services without ads, but it all reads like utopian fantasy stuff.

Do you genuinely believe the economy didn't function prior to the invention of the advertisement?

Yeah they probably had fewer advertisements back in the good old days, but they also had a much smaller economy, producing a whole lot less than we produce today. [1]

And you might not like everything about the world today, but living in a vibrant economy where people create wealth by building new businesses has led to more comfortable lives for the large majority of humanity.

[1] - https://ourworldindata.org/grapher/global-gdp-over-the-long-...

Post reply on HN