Hacker Newsnew | past | comments | ask | show | jobs | submit | mbrumlow's commentslogin

Decisions are less costly. When a swe can take 4 days to do what would have cost 6 months, the math of making sure you are doing the right thing before executing goes away.

That has little to do with building code and a lot to do with customers and releases and operations - giant companies don’t just magically demo software to people.

There’s so many layers today that can’t exist if this is the way forward.


> Now I'm pretty sure that people who say they "vibecoded an app in 30 minutes" are either building simple copies of existing projects, produce some buggy crap, or just farm engagement.

Some people seem to be better at it than others. I see a huge gulf in what people can do. Oddly there is a correlation between was a good engineer pre AI and can vibe code well.

But I see one odd thing. A subset of those who people would consider good or even amazing pre AI struggle. The best I can tell at this stage is because they lacked get int good results with unskilled workers in the past and just relied on their own skills to carry the project.

AI coders can do some amazing things. But at this stage you have to be careful about how you guide it down a path in the same way you did with junior engineers. I am not making a comparison to AI being junior, they by far can code better than most senior engineers, and have access to knowledge at lighting speed.


True.

> he started his second job as the human labor behind AI sex bots

This makes no sense. AI sex bots don’t need humans texting and role playing.

I don’t think whoever wrote this article understands what AI is.

Note the fist job of describing what was in the video seems like somebody building a AI dataset.


Jason certainly does [1]. The commercial bots seamlessly traverse between AI, auto-respond and human.

It's very much an ensemble method. That's why people pay for it over just downloading an abliterated model from hf with system prompt hacking. Go and try it, the SOTA of role-playing models still have a lot to be desired

[1] https://computerhistory.org/profile/jason-koebler/


> The commercial bots seamlessly traverse between AI, auto-respond and human. It's very much an ensemble method.

This seems unlikely to me, given it'd increase costs and the response times would make it obvious.

The messages presented in the original source appear to be people expecting to be talking to a real person, likely on a dating app. The relation to AI is only speculative, and mostly in the direction of "my messages may be used to train a chatbot to replace my job of deceiving people" - which is plausible.

> That's why people pay for it over just downloading an abliterated model from hf with system prompt hacking.

I'd assume convenience, fine-tuning, and using a larger model than it's feasible for most people to run locally.


He goes over his coverage on the latest 404 podcast.

I agree it seems like it's only this one man that's making the claims though.

I looked for other people reporting the thing online but I couldn't find it it all tracks back to this one guy that Jason interviewed

Fails the double source test


“Actually Indians” is a meme for a reason

https://ia.acs.org.au/article/2025/the-company-whose--ai--wa... The company whose ‘AI’ was actually 700 humans in India

https://www.businessinsider.com/amazons-just-walk-out-actual... Amazon's Just Walk Out technology relies on hundreds of workers in India watching you shop

https://www.cbsnews.com/amp/news/former-nate-ceo-human-worke... Former nate CEO who used human workers instead of AI allegedly defrauded investors lured by new tech of millions


(We've since changed the URL from https://www.404media.co/ai-is-african-intelligence-the-worke... to the essay it references, which was written by Michael Geoffrey Asia himself and goes much deeper into the topic. I've put a reference to the profile article in the toptext.)

[flagged]


You must not see Facebook ads. Every other one is for an "AI Girlfriend" no matter how many I hide.

You’re being targeted, presumably based on your traffic profile. Unless you’re talking about ads actually on Facebook, in which case your problem is that you’re using Facebook.

Yea I use facebook marketplace to sell stuf

Correct. Never on Facebook, don't have the app. On a browser I always use AdBlock.

No dude. Read it again.

The VC lends (the money from the bank) which the vc borrowed, to the clinic.

They are a sort of middle man. It the clinic is on the hook to the bank and the Vc takes fist cut before playing the bank.

Eg. The vc only risked the company they were buying, and gets paid first.


If the VC borrows money from the bank and lends it to the clinic, the clinic is not on the hook to the bank. The clinic is on the hook to the VC and the VC is on the hook to the bank. Which means that if the clinic goes under, the VC takes the loss because it still has to repay the bank.

(Edit: To be clear, I agree with the other commenters that none of this is what VCs do. I'm just pointing out that the way this is being described doesn't even work on its own terms. Needless to say, LBOs are not "risk free".)


Nope. The clinic is the collateral to the bank. VC stand to loose nothing.

It does not happen overnight. But what happens is after they take control of the clinic or company they change the sales model to boost reoccurring revenue, this then allows the clinic or target company to take loans out. Because they look good on paper. The company then pays VC back when then pays bank back.

This can be done in about 6mo to 1 year process with some companies. The initial out of pocket expense is small and paid back very quickly.

I also forgot. Sometimes they will take the newly owned company and merge it. During that process they extract more money and load more debt onto the remaining entities, again making the VC money.

In some cases they can even get huge tax benefits by loading the company with debt which offsets the tax bill of the final entity.

When these transactions are done, within the span of a day multiple companies are created and merged and absolved.

There is little to no risk for the VC


> The clinic is the collateral to the bank. VC stand to loose nothing

This is actually a case where using the correct terminology clarifies.

VCs don’t do LBOs. Private equity firms do. When their deals go bust they lose the equity they invested. That equity is the first layer to take a loss. When that happens, the lenders—whether they be banks or private credit firms—take over the company, often converting some of their previous debt into equity.

There is a lot of risk in LBOs. It’s why they have such a mixed record.


PE includes buy-out (leveraged and not) and VC transactions. PE is typically any medium to long term equity investment not traded publicly on an exchange. Even this is cloudy now that the PE firms themselves are going public.

LBOs are also not a black and white classification, at least not the way they were in the Gordon Gecko 80's, with varying levels of target-borne debt financing specific to the deal. So while I agree "VCs don't do LBOs", PE does both LBOs and VC deals, with the PE firms doing their own style of fund and deals.

I found this book (though dated) to be a more academic analysis of PE: https://www.wiley.com/en-us/Private+Equity%3A+History%2C+Gov...

don't buy it; try your local library.


> this is cloudy now that the PE firms themselves are going public

The public or private status of the manager has no relation to private equity being cloudy. Out of all of the delineations, PE is a pretty sharp one. VC is PE. Private credit is not. It’s private debt. Not equity.


> It does not happen overnight. But what happens is after they take control of the clinic or company they change the sales model to boost reoccurring revenue, this then allows the clinic or target company to take loans out. Because they look good on paper. The company then pays VC back when then pays bank back.

This was the missing bit for me. Thanks for taking the time to explain!


Who are the bagholders in these scenarios?

The PE sales pitch is often that the target company can benefit from expertise management and/or there is value locked in it that can be captured. Both of these claims are... marginal? Studies around the "expert management" claim tend to show this is not true, based on pre/post returns, but it's hard to account for the long term, because PE also tends to focus on sales with very specific characteristics & time horizons (and associated cost savings) that benefit a 5-7 year fund that sells the portfolio company (wait for it) around years 3-5.

Which is a long-winded way of saying the bag holders are anyone invested in the long-term success of the company: 1. employees, 2. customers, 3. owners (i.e. the next PE fund) when the music stops, i.e. what we saw when interest rates went up impacting debt financing, and (real or not) AI-eats-SaaS impacted valuations. I'll add 4. "the public" if the company is big enough, with various levels of goverment and employment, taxes, etc. lost but I think it's more the smaller organizations in aggregate that hurt at this level than any specific company.


What's the betting that it's (somehow, eventually) the taxpayers?

> If you actually want to engineer properly.

I think this statement is misguided, and potentially comes from a lack of experience in getting AI coders to produce quality.

Proper engineering does not come about from the tools you use or how you use them. Proper engineering has always come from thought, and reasoning, it never was about the act of coding. It always was about the systems thinking and expressing the goals and desires that matched the requirements.

IDEs were never needed to properly engineer and in the days of AI will become increasingly less important.

Tools for planning, reviewing, and commenting on code are the future. The necessity to edit actual code is coming to an end.


Yes, that's what I said, I'm contrasting properly engineered AI code to vibe coded slop AI code, not that human written code is inherently better engineered.


The they failed part is not the Senate. Some members of the Senate failed to get the votes to pass the bill. Whe. You say it like you have it implies the will of the senate was to block it and they failed. And that is weird because the senate clearly did not want to block it otherwise it would have been blocked.

The senate rejected the bill. They did not fail to do anything. The bill failed to get the approval of the senate.

But one thing we know. In the senate did not fail.


Something about the wording seems dishonest though. Whoever sponsored the bill failed to get the senate to pass it.

The wording here makes it seem like the senate wanted this but failed to get it.

So again the senate failing to do something the senate said they did not want to do is weird.

It comes down to some people in the senate wanted this, but they are not the senate.

Politicians have been treating a minority position as the institution’s will for some time. It’s our job to look past that and not be fooled even if you share the same minority position.


> author has very strong opinions.

So true. To the point I have to maintain my own fork to make the command key my meta


GIAC has zero authority, any group of people can get together and make their own policies and print a nice little certificate when somebody applies.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: