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

73% of AI startups are just prompt engineering

pub.towardsai.net

Post reply on HN
Print view View original

201–210 of 212 posts

Re: 73% of AI startups are just prompt engineering

#201

Post by s1mplicissimus »

apwell23 wrote: post #167
scuff3d wrote: post #166

Earlier quoted context omitted.

I can't take anyone seriously who uses prompt engineering unironically. I see those emails come through at work and all I can do is roll my eyes and move on

what level of seriousness does "context engineering" deserve?

Assuming you mean "context engineering" as in "engineering the context for LLM prompts" - 0.

I take particular issue with the usage of the word "engineering", in this context, as in my practical experience what I witnessed was more akin to "try random things until it somewhat works" than anything I would associate with "engineering". But hey, it's a free country and people can use words whichever way they want. Just shouldn't be confused if noone keeps on listening ;)

Re: 73% of AI startups are just prompt engineering

#202

Post by FridgeSeal »

zkmon wrote: post #196
nottorp wrote: post #195

Earlier quoted context omitted.

Great analogy. I don't eat at those kinds of "restaurants" though. If I buy pre made it's at best from supermarkets for heating at home.

If restaurants can't use pre-made food, they would have to have a farm in their backyard with all sorts of animals, veggies and grain crops.

What an incredible false dichotomy.

Yes clearly the only 2 options are “sells someone else pre-made meals” or “farms everything on premises”.

Clearly no other alternatives exist. /s

Re: 73% of AI startups are just prompt engineering

#203

Post by mvkel »

kilroy123 wrote: post #65
mvkel wrote: post #41

Wait til you hear what GPT 5 is

What is it? A gpt-4o wrapper?

It's not a new model; it's a router + many, many chains of optimized system prompts. Note the terminology they use to describe it.

> We are introducing GPT‑5, our best AI system yet

> GPT‑5 is a unified system with a smart, efficient model that answers most questions...

They make a special point to call it a system, of which "a model" is a part. I'm betting they didn't pre-train gpt5 from scratch, but did a ton of RL on gpt4 + prompt optimization and the routing engine.

Re: 73% of AI startups are just prompt engineering

#204

Post by hotpaper75 »

yalogin wrote: post #109
hotpaper75 wrote: post #71

Earlier quoted context omitted.

Do not forget that a product idea needs to meet a certain ROI to be stolen. Big Tech won't go after opportunities that do not generate billion-level revenue. This leaves some room for applications where you can earn decent money.

That is not how companies work. What you said may be true for the immediate short term but over time every team in the company needs to show improvement and set yearly milestones. All these startups will then become functionality they want to push that quarter. Yes it doesn’t mean the death of the startup but a struggle

Sure but at the end I believe one of the deciding factors on which projects you pursue is the ROI. In a scenario where a given team has many projects it can go after it has to select which project to chose from.

Re: 73% of AI startups are just prompt engineering

#205

Post by ankit219 »

zkmon wrote: post #84

But ... what else should they be doing? What's the expectation here? For example, in the 90's, a startup that offered a nice UI for a legacy console based system, would have been a great idea. What's wrong with that?

Maybe, not lie about what they are doing? That is not unreasonable as a base expectation.

Re: 73% of AI startups are just prompt engineering

#206

Post by xela79 »

and about 90% of consumers of these AI startups have no idea how to prompt engineer.

at this time that just translates to "73% of software startups are just writing code in C++"

yes you could write your own code, your own prompt for LLM, or you could spend time on using the code / LLM output

Re: 73% of AI startups are just prompt engineering

#208

Post by RC_ITR »

bradfa wrote: post #58
RC_ITR wrote: post #17

One of the biggest problems frontier models will face going forward is how many tasks require expertise that cannot be achieved through Internet-scale pre-training. Any reasonably informed person realizes that most AI start-ups looking to solve this are not trying to create their own pre-trained models from scratch (they will almost always lose to the hyperscale models). A pragmatic person realizes that they're not f…

Got any links to explanations of why fine tuning open models isn’t a productive solution? Besides renting the GPU time, what other downsides exist on today’s SOTA open models for doing this?

When the new pre-trained parameters come out in a new model generation, your old fine tuning doesn't apply to them.

Re: 73% of AI startups are just prompt engineering

#209

Post by hn_throwaway_99 »

chmod775 wrote: post #192
teiferer wrote: post #147

Earlier quoted context omitted.

> Monitored their network traffic for 60-second sessions How can he monitor what's going on between a startup's backend and OpenAI's server? > The truth is just an F12 away That's just not how this works. You can see the network traffic between your browser and some service. In 12 cases that was OpenAI or similar. Fine. But that's not 73%. What about the rest? He literally has a diagram claiming that the startups con…

> How can he monitor what's going on between a startup's backend and OpenAI's server? He is not claiming to be doing that. He says what and how he's capturing multiple times. He says he's capturing what's happening in browser sessions. Reflect on what else you may to re-evaluate or discard if you misunderstood this. > That's just not how this works. You can see the network traffic between your browser and some servic…

> He is not claiming to be doing that.

He literally is. Look at the sequence diagram in the article, which specifically labels "Startup Backend". The whole article is made-up LLM slop.

Re: 73% of AI startups are just prompt engineering

#210

Post by harambae »

yawnxyz wrote: post #119

it's wild, I work with some fortune 500 engineers who don't spend a lot of time prompting AI, and just a quick few prompts like "output your code in ... " tags" — a trick that most people in the prompting world are very familiar with, but outside of the bubble virtually no one knows about — can improve AI code generation outputs to almost 100%. It doesn't have to be this way and it won't be this way forever, but this…

Can you clarify what this XML tag formatting does?

I use Claude Code and Gemini CLI pretty extensively and haven't come across this yet.

Post reply on HN