Ask HN: Why is every company encorporating AI everywhere?

As someone who has experience building a small AI powered company and making these decisions on a small scale, I can say maybe?

For some companies, automation that is based on being able to pull info from a gigantic knowledge base that still has bounds is helpful for them. Most orgs are still figuring this out in the face of immense pressure from their stakeholders not to fall behind and miss any potential windfall. They don’t know how LLMs can help them but the hype may push them in the black or lead to (false) growth so they lean in.

I think many of the larger tech companies were hoping to make human engineers obsolete so that they could achieve the true promise of SaaS - almost no investment, instant money faucet. Unfortunately for them, we still need people to help us understand people so engineers, designers, and product managers aren’t as expendable as some of them were hoping.

No one thought AI would be free or cheap, but a LOT of ppl bet that the true cost could be obfuscated or waved away temporarily while the money was stupid and then revealed once enterprise was addicted and struggled to manage their switching costs. By then the smartest folks have made their money, moved it to safer havens, and await the next stage of the lifecycle.

I love LLMs but the way we use them seems very stupid, IMHO. It’s like everything is a screw now but everyone still has bolts and nails.

Because VC money depends on the company’s growth plans, and VC money isn’t offered if “AI” isn’t part of the product any more. (Sample size of one, told by a friend who’s in direct contact with the CEO.)

From the POV of a CEO(not myself), it's a predictably priced, affordable insurance against potentially falling behind the competition.

If the AI is deeply integrated and feature-rich, I would prefer it. However, if it is just a disconnected chat interface, I don't think it is necessary. Additionally, whether it requires an extra fee is a factor that must be considered.

Related: https://news.ycombinator.com/item?id=48964185

If enough people/companies are on the bandwagon, you're afraid to push back and be the crazy one, potentially losing money. So you roll with it.

I've found the new AI customer support bots that a lot of websites started using are usually more knowledgeable and honest about whatever product they represent than human agents are. I'm sure that will change eventually when they start figuring out how to lock down the AI to a "safe script" the same way they train human customer support, but for now, it's pretty great.

:-) I have had my share of bad meetings over this, I have always advocated to start with enhancing existing features of product that would benefit from AI. But the org mandate is compulsory one idea for each product and new products should be agentic or RAG based only. I don't understand this rush to incorporate AI, I feel shit must be organic and self paced, forced adoptions wont last long.

Someone I know said it's because investors get excited to know your product has AI.

I think for a good share of products it isn't necessary, but I also guess that it's very hard to see from the start. I strongly believe there generally is value in AI, but finding this particular application that unlocks it -- is a conundrum. So, if I imagine myself being a head/vp of product, in one of these companies, I have pretty much no choice but to integrate it and study the metrics. Then, after a few iterations the value is either found or not, and depending on the sanity of stakeholders, the useless AI features can be retired. But not before the evidence is collected, because you can't fight the AI frenzy without ones, lol

Product owners were promised that LLMs/AI is a magical genie that can do anything, so they make it do everything. The trillions of dollars spent on data centers and thousands of layoffs has to be justified somehow.

I don't mind AI being added to products, I just want it to solve a real problem instead of becoming another button that I never click.

FOMO.

There are most likely many different reasons. Cutting costs is a big one; leveraging new functionality in a simpler way is another one. For instance, yesterday I watched some youtube videos and one use case was "count to 100 in a very sad voice", even if that is more a joke and ... pretty useless on its own. The AI did.

What was interesting to me here was that if I were to write a program to do the same, I could most likely do that (just simplify the tasks as they are and implement each one), but it would take me longer. As much as I hate AI, there are real benefits in at the least some areas.