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

95% of Companies See 'Zero Return' on $30B Generative AI Spend

thedailyadda.com

Post reply on HN
Print view View original

401–410 of 445 posts

Re: 95% of Companies See 'Zero Return' on $30B Generative AI Spend

#401

Post by dvfjsdhgfv »

jawns wrote: post #56

Full disclosure: I'm currently in a leadership role on an AI engineering team, so it's in my best interest for AI to be perceived as driving value. Here's a relatively straightforward application of AI that is set to save my company millions of dollars annually. We operate large call centers, and agents were previously spending 3-5 minutes after each call writing manual summaries of the calls. We recently switched to…

What I found is that although some tools are good for transcription and the final result is quite accurate, this is not necessarily true for summaries.

Yes, I routinely read meeting summaries spit out by the so-called SOTA tools and they almost always contain major errors. So if you actually need them for anything serious, it would be wise to keep the audio.

Re: 95% of Companies See 'Zero Return' on $30B Generative AI Spend

#402

Post by rawgabbit »

resiros wrote: post #224

Here is the report: https://mlq.ai/media/quarterly_decks/v0.1_State_of_AI_in_Bus... The story there is very different than what's in the article. Some infos: - 50% of the budgets (the one that fails) went to marketing and sales - the authors still see that AI would offer automation equaling $2.3 trillion in labor value affecting 39 million positions - top barriers for failure is Unwillingness to adopt new tools, Lack…

This stood out to me in the report:

      A corporate lawyer at a mid-sized firm exemplified this dynamic. Her organization invested $50,000 in a specialized contract analysis tool, yet she consistently defaulted to ChatGPT for drafting work: "Our purchased AI tool provided rigid summaries with limited customization options. With ChatGPT, I can guide the conversation and iterate until I get exactly what I need. The fundamental quality difference is noticeable, ChatGPT consistently produces better outputs, even though our vendor claims to use the same underlying technology." This pattern suggests that a $20-per-month general-purpose tool often outperforms bespoke enterprise systems costing orders of magnitude more, at least in terms of immediate usability and user satisfaction. This paradox exemplifies why most organizations remain on the wrong side of the GenAI Divide.

Re: 95% of Companies See 'Zero Return' on $30B Generative AI Spend

#403

Post by dvfjsdhgfv »

dsr_ wrote: post #221
andix wrote: post #201

Earlier quoted context omitted.

Imagine a follow-up call of a customer. They are referring to earlier calls and the call center agents needs to check what it was about. So they can skim/read the transcripts while talking to the customer. I guess it's really hard to listen to transcripts while you're on the phone.

That would be awesome! But in fact, customer call centers tend not to be able to even know that you called in yesterday, three days ago and last week. This is why email-ticketing call centers are vastly superior.

I'm in love with email-based support as I am on both sides of the chain. When I raise a problem, the engineers on the other side can work at their pace, escalate when needed, and I almost always get a reasonably good reply. I can dig deeper if I wish, and I'm pretty sure the guys on the other end are doing their best.

It works the same way when I'm helping someone else: most reasonable people don't expect that if they make an audio call I will magically solve their problem faster. Maybe it will be slower and they will get a lower-quality ad-hoc solution.

Re: 95% of Companies See 'Zero Return' on $30B Generative AI Spend

#404

Post by lacy_tinpot »

yifanl wrote: post #256
lacy_tinpot wrote: post #226

Earlier quoted context omitted.

It doesn't because we already know it doesn't because just like think about it and it clearly doesn't. So everyone trying to make the technology is a big dumb dumb for even trying. Like is the conclusion we shouldn't even try? This kind of thinking ridiculous.

Same reason why we aren't looking for innovative ways of using sugary soft drinks as a building material. Just because there's a non-zero chance we could come up with something isn't compelling enough by itself.

Your analogy serves zero purpose.

What's the logic here? Machines can't think because humans are the only ones that can think. We know because we're thinking and the machines can't.

So what's thinking? Well we're doing it and it's obvious.

That's just bad thinking. Worse than these machines actually.

Re: 95% of Companies See 'Zero Return' on $30B Generative AI Spend

#405

Post by deadbabe »

morkalork wrote: post #297
deadbabe wrote: post #261

Earlier quoted context omitted.

What’s the ROI?

Transcription cost is a race to the bottom because there's so many vendors competing, same with embeddings. It's positive. Gets better every year.

So no ROI, just hope.

Re: 95% of Companies See 'Zero Return' on $30B Generative AI Spend

#406

Post by paulddraper »

Spivak wrote: post #174
paulddraper wrote: post #141

Earlier quoted context omitted.

This works unless you want to automate something with the transcripts, stats, feedback.

Why wouldn't it, once you actually have that project you have the raw audio to generate the transcripts. Only spend the money at the last second when you know you need it. Edit: Tell me more how preemptively spending five figures to transcribe and summarize calls in case you might want to do some "data engineering" on it later is a sound business decision. What if the model is cheaper down the road? YAGNI.

The "last second" is right after the call.

For example, if 60% of your calls this month mention assembly issues with the product, that information will help you improve it.

This is practical, not theoretical.

Re: 95% of Companies See 'Zero Return' on $30B Generative AI Spend

#407

Post by a-posteriori »

AstroBen wrote: post #8

The link to their referenced study doesn't seem to work? This is confusing.. it's directly saying AI is improving employee productivity, but that's not leading to more business profit... how does that happen?

This study was led by a PI in the MIT Media Lab who studies cameras and their impact on society.

There's a reason it has a catchy headline, there's a reason you needed to fill in an email form to get access to the study, and there's a reason why the 2nd author has an agentic AI startup.

I thought it was a low quality article with no data or in detail methods. MIT needs to do better. This is the second article of this type in the last few months that caught headlines.

Re: 95% of Companies See 'Zero Return' on $30B Generative AI Spend

#408

Post by bmicraft »

josefx wrote: post #335
Jolter wrote: post #260

Earlier quoted context omitted.

Perhaps doing this suggested auto-summarizing would be what finally solves that problem?

Is doing that going to be cheaper than not doing it?

Maybe, if it means people spend less time on calls (because their problem got solved sooner?)

Re: 95% of Companies See 'Zero Return' on $30B Generative AI Spend

#409

Post by Jolter »

ghalvatzakis wrote: post #373
Jolter wrote: post #317

Earlier quoted context omitted.

What kind of interviews do HR do, apart from job interviews?

Not HR either. I work for an experts network firm

Aha, see, your choice of words there was ”freed human resources” which threw me off. I see now.

Re: 95% of Companies See 'Zero Return' on $30B Generative AI Spend

#410

Post by morkalork »

ch4s3 wrote: post #398
krainboltgreene wrote: post #382

Earlier quoted context omitted.

That's what everyone says they'll do and then it never gets touched again.

I guess you just know better than everyone, include the people who do look at user interactions. I know I've done it, so I must be no one.

I guess I'm no one too because I've done plenty of call analyses too.
Post reply on HN