Proof of Care in the Age of AI – Jacob Filipp’s Handwritten Blog Experiment

Proof of Care in the Age of AI – Jacob Filipp’s Handwritten Blog Experiment

The core idea: hand‑written content as proof of care

Jacob Filipp created a blog post that is literally handwritten, photographed, and then turned into a copy‑pasteable web page. The effort demonstrates that the author cared enough to invest time and manual labor, a signal that is hard for large language models to replicate convincingly.

How the post was built – step‑by‑step technical walk‑through

1. Hand‑write, photograph, and assemble

Filipp wrote the entire essay on paper, took high‑resolution photos of each page, and combined the images into a single SVG file. This keeps the visual authenticity of the handwriting while allowing web‑based rendering.

2. Add invisible spacing for clean copy‑paste

To make the text selectable, Filipp overlaid the SVG with a text layer that mirrors the handwritten words. He inserted two kinds of invisible characters:

  • ` (backticks) as narrow spacers that are removed entirely by JavaScript.
  • Multiple regular spaces that are collapsed to a single space during the copy event. These characters ensure that when a reader copies the text, they receive clean, normally spaced prose.

3. Create hyperlinks directly in the SVG

Using Inkscape’s Create Anchor feature, Filipp added clickable links to the overlaid text. This preserves the interactive experience of a normal HTML article while keeping the handwritten aesthetic.

4. Post‑process the SVG

After exporting, Filipp performed three essential modifications:

  1. Embed the font – a <style> block embeds Barlow Condensed – Italic Condensed, matching the dimensions of his handwriting.
  2. Replace backticks – backticks are swapped for six‑per‑em space characters (U+2006) so they remain invisible when selected.
  3. Convert local image paths to absolute URLs – this ensures the SVG loads correctly on any host.

5. Inline the SVG in HTML

Instead of embedding the SVG via an <object> tag (which would prevent JavaScript copy‑event handling), Filipp inserted the <svg> markup directly into the page’s HTML. This enables the custom copy‑paste script to run when users select text.

Why this matters in an AI‑dominated content landscape

Human effort as a credibility signal

When large language models can generate plausible prose in seconds, readers lose a reliable way to tell whether a piece reflects genuine human thought. By requiring a labor‑intensive process—handwriting, photographing, and custom SVG processing—Filipp provides a proof of care that is difficult for AI to fake.

Potential applications beyond blogging

  • Academic submissions – a handwritten supplement could verify original work.
  • Code reviews – a “human‑only” explanation of a patch could be required before merging.
  • Online identity verification – services could ask users to submit a short handwritten note to prove they are not bots.

Community reactions – insights from Hacker News comments

"Reading this made my day, not just because of the content, but because someone else cared enough to tackle the same problem." – malty_on_rock

"If I had handwritten this, there would be at least one (likely lots more) errors… It’s plausible that the author cleaned up the text after photographing." – jimmiles

"Typewriters work too. I’m toying with the idea of an online journal that only accepts submissions written by hand or by mechanical typewriters." – theopsimist

"Proof of care still leaves the audience to determine if the care is in the message or in attracting attention." – nlawalker

"I read the reflected‑letter text and found that amusing." – wolttam

"Handwriting forces you to think more, but it isn’t a guaranteed proof of humanity; AI can generate realistic handwriting images." – halfax

"A vlog with one‑shot recordings could be another ‘farm‑to‑table’ approach to proving effort." – bearjaws

"The idea maps well to software development: require a human‑only explanation of a PR before it can be merged." – gnarlouse

"Proof of care may be valuable, but the reading experience suffers if the handwriting is hard to read." – voidUpdate

These comments highlight both enthusiasm for the concept and skepticism about its practicality and readability.

Limitations and open questions

  • Readability – Handwritten text can be illegible for many readers, reducing the utility of the content.
  • Scalability – The workflow is time‑consuming; it may not suit high‑volume publishing.
  • AI imitation – Advances in generative models can synthesize realistic handwriting, potentially eroding the uniqueness of this proof.
  • Accessibility – Screen readers cannot interpret images of handwriting without OCR, which may be unavailable for custom scripts.

Takeaways for creators and platforms

  1. Signal effort deliberately – Whether through handwriting, video recordings, or other labor‑intensive media, explicit effort can differentiate human‑generated content.
  2. Balance authenticity with usability – Provide a clean, selectable version (as Filipp does) to avoid alienating readers.
  3. Consider alternative proofs – Time‑stamped drafts, version histories, or cryptographic commitments can also demonstrate care without sacrificing accessibility.
  4. Prepare for AI countermeasures – As AI handwriting generation improves, combine multiple signals (e.g., unique personal anecdotes, metadata) to maintain credibility.

Conclusion

Jacob Filipp’s handwritten‑to‑SVG experiment is a concrete demonstration of proof of care: a tangible, labor‑intensive process that signals human involvement in a world where AI can produce text instantly. While the method raises practical concerns about readability and scalability, it sparks a broader conversation about how creators can credibly convey effort and authenticity in digital media.

Sources