[pyman]

Thanks for sharing this inside information. It's really helpful and gives me (and others researching this case) a better understanding of what was happening behind closed doors.

I have a couple of questions, if you don't mind:

1. I checked on LinkedIn and saw that most employees are based in India. Is that correct?

2. From what I understand, customers knew their apps were being built by developers, and they could use an internal tool called Studio to track progress and view the names of the developers assigned to tasks. Is that accurate?

3. Did you work with Craig Saunders, the Head of AI? I heard he was attending events and demoing some internal apps he had built, and that people were pretty impressed. Do you know what exactly he was showing?

4. The builder.ai website never claimed to use GenAI, only AI, and it clearly says that their virtual assistant, Natasha, assigns developers to projects. Have you ever seen Natasha doing this in action?

Thanks again for speaking up and helping clear up some of the confusion.

[redeyedtreefrog]

My inside information is both outdated and minimal, I was there for a short time a few years ago.

As far as I know all the development work on client projects was done by developers in India. I was too far removed from the details to know quite how they were employed and paid - I literally never had any interaction with them - but I suspect this is the main source of the LinkedIn employees.

As far as I know customers were aware apps were being built by human developers, the "80% by AI" was prominent on the website and the sign up process but it was never claimed the whole thing was automated.

It looks like Craig Saunders joined only a year ago, in June 2024, which is long after I left. Given how much AI has advanced over the last few years, and given his LinkedIn shows he previously held a senior AI role at Amazon, I wouldn't be remotely surprised if he and other employees built some impressive tech demos over the last year or two.

Regarding Natasha assigning developers to projects, I don't know to what extent it was really automated, I wasn't involved in that.

[pyman]

Thanks for clarifying that.

Here's the thing. The website doesn't say that 80% of the code was created using AI. That's what the fake article claims. What the site actually says is:

---

Natasha is your AI product manager.

Using machine learning algorithms, she recommends the features you need, based on the type of app you're building.

Natasha also creates an instant prototype for you, helping visualise your idea.

80% of this information is gathered automatically.

---

A few things to point out:

1. The site says it uses algorithms, not GenAI, to make recommendations and create a prototype. I assume that means some visual mockups or images of the suggested features.

2. It says it creates "80% of the information", not 80% of the code. That line is clearly referring to the recommendations, not the final software.

People spreading fake news scanned the site, misunderstood how AI works, and twisted the wording to fit their own narrative.

[redeyedtreefrog]

If you scroll down it also says 60% of the code is automated. Both percentages are essentially meaningless. If you require customers to type all their requirements into a static web based form does that also count as 80% automated information gathering? If 60% of an application is templated components does that mean code generation is 60% automated?

It's meaningless, and I suspect intentionally so. It clearly implies that they were doing things in a more advanced way than a standard Indian outsourcing firm, an implication heightened by hiring people to build impressive tech demos. But the impressive tech demos and the work on real client projects were two entirely separate things.

[pyman]

I agree, it's confusing. It looks like marketing was throwing numbers around without really explaining what they meant.

Automation is definitely not AI. Maybe the 60% referred to reusable code, or a theme, or a template that was handed off to the developers?

Either way, 60% feels like a random number.

[pyman]

Small correction: automation combined with machine learning is considered AI.

So the 60% was probably referring to the amount of code (templates, components, etc) they were reusing.