[piker]

This is great advice (that we need to follow) but needs to be updated for 2026. The information value of providing (or receiving) a demo has dropped to roughly zero with vibe coding. Today, an apparently functional and useful product can be produced and demoed in minutes, but that demo provides absolutely zero information into the technical capabilities of the demoing team to follow through on promises with polish and at scale. It doesn't reflect a studied architecture or edge case handling. It basically only shows a vision, which can be tailored to perfectly mirror the recipient's expressed desire even though it's absolute vaporware. This makes it even harder to sell to enterprise in 2026 when the scene is awash in such noise.

[operatingthetan]

>that demo provides absolutely zero information into the technical capabilities of the demoing team to follow through on promises with polish and at scale.

With vibe coding comes vibes-based capital. I'm only half kidding.

[QuantumGood]

Speed to market has always been a factor. Venture prioritizing this factor due to AI accelarating speed is probably expanding ... for now. In bubbles, speed tends to rise to a higher priority.

Yes, first/fast is sometimes a negative factor (e.g. first to market doesn't mean best, second to market can take advantage of proof of market proviced by first, etc.)

[keiferski]

In my experience demos are half about the product and half about the team / company behind it. So I wouldn’t call its value zero: part of the reason a potential client is asking for a demo is to see if there’s actually a real, intelligent company behind the product.

[bonesss]

The sales pitch needs to compete with other pitches, so I gotta imagine in a vibe-heavy market a solid sales team is gonna lead with all the stuff you can’t vibe.

Customer transaction numbers, service response times, human staffing for VIP customer service, and human engineers who are recognized domain experts. The cliche live call to customer support with some hairy-ass customer specific problem.

Plus vibe-upselling of vibe-integrations for whatever Wonderful Engineering the customer has with your profit centres.

[oliver236]

AI demos are 5% of the product

[stogot]

I’ve seen AI demos that are 100% the product, and the “company” won’t renew the $10 domain at the end of the year

[appplication]

No offense but that is a wild statement

[cmrdporcupine]

Right, and the story now shifts to: What's your customer service & support model? How can you prove this is stable and that you can maintain it? Who is going to handle the pages in the middle of the night?

[mlnj]

All those things are beyond the demo itself. Vibe-coded demos are just demos. There are stability, security and everything enterprise that still needs to be added to a demo to actually make it functional as a paid offering.

The hard problems still remain.

[cmrdporcupine]

My point is that if I were an investor in the LLM-era I'd be shifting my attention to the answers to those questions more than I would to tech demo.

[user_7832]

> The information value of providing (or receiving) a demo has dropped to roughly zero with vibe coding.

Only if you're a software-only startup. If you have hardware, the entire article is still valid.

[bjornroberg]

The artifact can be faked cheaply now, so the only buying signal left is commitment. That's exactly the "ruthless" move the post argues for, I think.

[eiiejr]

Vision and strategy are more important than ever.

And frankly visually being able to explain how your product beats the competition is more important than writing lines of code for a product that could be DOA.

However not everyone can do this. So the scientific approach gets pushed.