The X I Hoped For

A conversation with Grok on Truth, Justice, and AI

When Elon Musk bought Twitter in 2022, promising a freer platform rooted in free speech, I was excited for the future of social media. As someone who was permanently banned pre-Elon for daily reminders to Ted Cruz about Uvalde and gun safety — with no chance of appeal, and no recourse — I saw hope.

[embed]Twitter at it again. Suspended with no reason given as to why…medium.com

Musk’s vision of “Freedom of Speech, Not Freedom of Reach” resonated with my belief that justice, not just law and order, should guide discourse.

The old Twitter silenced me for pushing a truth: it was time to discuss gun safety post-Uvalde. No violence, just persistence. The ban felt unjust, a tool of censorship.

When Elon Musk took over, I rejoined, excited by his AI-driven, truth-seeking ethos. I contribute to writing and voting on Community Notes, fact-checking posts to uphold accuracy. But the platform isn’t the revolution I envisioned.

Recent experiences with a shadow ban for a fact-based tweet, have left me disappointed. In a recent chat with Grok, xAI’s curious AI, we explored how X could fulfill that promise — and prevent a darker future.

Recently, I tweeted in Danish:

Translated it says: How many people in Denmark know that Danes returning from abroad receive less than half the support that asylum seekers and so-called refugees get from the state? Yes, Danes are treated worse than immigrants.

That offended someone, and someone posted this:

Translated it says: Your post was marked as offensive by X. I wonder if they borrowed Facebook’s censors?

What I wrote was a fact (per uim.dk, skat.dk): Returning Danes get around 6,000 DKK, while asylum seekers/refugees get 12,000 DKK.

No attacks on any religion, culture, or ethnicity — just data. Yet, it was reported as offensive, and my interaction dropped to zero — a shadow ban.

Meanwhile, accounts with hundres of thousands followers post “death to the IDF” unchecked. The inconsistency is staggering.

I turned to Grok, xAI’s AI, on July 4, 2025. We dissected this. Grok agreed my tweet was factual and saw no attack, yet noted X’s safety-first algorithm flagged it due to immigration sensitivity. We contrasted this with high-follower hate speech, suggesting larger accounts evade scrutiny due to scale and algorithmic limits (Social Media Today, USC Viterbi).

Grok supported my shock, aligning with xAI’s truth mission.

I proposed solutions: Using AI to auto-flag debunked reposts (building on Ars Technica’s Community Notes experiments) and track record-based moderation — rating users by lie frequency (Notes rejections) and violation severity.

Grok endorsed this, suggesting a trust score (e.g., 90% accuracy) to deter bans on truth-tellers like me. We agreed bans should be harder unless intent is clear, like “death to” calls.

The Dark Future We Must Prevent.

Europe’s trend alarms me. Police forces, funded by EU programs (CORDIS, 2012), now monitor social media, using comments to justify force — e.g., German hate speech arrests (dw.com, 2023–2025).

The DSA amplifies this. I fear a future where truth is policed, not protected.

[embed]The EU's Digital Services Act A common set of EU rules that helps better protect users' rights online, bring clarity to digital service providers and…commission.europa.eu

Grok and I believe AI can counter this, ensuring facts prevail over tyranny. Justice, not order, is what we crave.

The X I Want and Envision.

X should be a beacon of truth. Auto-flagging lies, tracking reliability, and easing bans on factual posts would fulfill Musk’s vision. I’ve asked Grok to help draft this pitch, and I’ll test it via Notes feedback.

Community, join with #XVision — share your stories. Tag @ElonMusk and @XEng, let’s co-create the future of X and Free Speech worldwide.