Experiment liveDay 48 / No filters

Independent field experiment

fuckthealgos.tech

Remove your filters.
Connect with the human layer.

Tracking the algorithmic fallout of unfiltered communication on LinkedIn.

View live data
Section 01

Fuck it.

Run the experiment. Watch what changes when the polished version stays in drafts.

Recorded to date: 4,266 impressions // 125 engagements // Peak day 10/12: 280 impressions // Followers 1,438 · zero net loss // 8 weekly exports logged // Raw data open on GitHub    //    Recorded to date: 4,266 impressions // 125 engagements // Peak day 10/12: 280 impressions // Followers 1,438 · zero net loss // 8 weekly exports logged // Raw data open on GitHub

01 / Algo reaction

Post impressions

313

↓ 40% vs. Oct 24–30

Members reached

167

Unique humans · Oct 31 – Nov 6

LinkedIn export · Oct 31 – Nov 6Updated: Nov 6, 2026Raw data ↗

Total followers

1,438

+2 net vs. prior week · 3 new this week

Engagements

4

Reactions + comments + shares · Oct 31 – Nov 6

Daily impressions + engagements · Sep 17 – Nov 6

Hover or tap a point for the readout. ━ impressions · ┅ engagements ×10

  • [W1] Sep 17–23 · Novelty trap: 865 impressions, peak 260 on 9/21.
  • [W2–W3] Sep 24–Oct 2 · Quarantine: reach fell to 309, up to −64% from peak week.
  • [W4] Oct 3–9 · Keyword rebound: 595 impressions, engagement flat at 18.
  • [W5] Oct 10–16 · Widest reach since launch: 858 impressions to 423 members; 17 engagements.
  • [W6] Oct 17–23 · Reach contracted: 482 impressions, 169 members, 9 engagements; followers grew by 13 net.
  • [W7] Oct 24–30 · Reach widened again: 520 impressions to 248 members; 9 engagements; followers up to 1,436.
  • [W8] Oct 31–Nov 6 · Quietest week yet: 313 impressions to 167 members; 4 engagements; followers 1,438.

Recorded to date · Sep 17 – Nov 6

4,266Impressions
125Engagements

Unique daily records; overlapping export dates counted once. Weekly reach cannot be added into a unique multi-week total.

Weekly record

PeriodViewsReachedEng.Followers
Sep 17–23865314371,418
Sep 24–30475184231,419
Sep 26–Oct 2309145171,419
Oct 3–9595267181,420
Oct 10–16858423171,421
Oct 17–2348216991,434
Oct 24–3052024891,436
Oct 31–Nov 631316741,438

New followers: +8 / +4 / +4 / +5 / +2 / +20 / +6 / +3 respectively. Totals are end-of-week snapshots.

Daily record · Impressions / engagements

9/1753 / 1
9/1866 / 2
9/19100 / 3
9/2068 / 5
9/21260 / 17
9/22198 / 7
9/23120 / 2
9/24159 / 7
9/25126 / 6
9/2671 / 7
9/2718 / 1
9/2852 / 1
9/2948 / 1
9/3044 / 1
10/150 / 4
10/226 / 1
10/332 / 0
10/439 / 1
10/546 / 2
10/6124 / 0
10/7117 / 3
10/8110 / 4
10/9127 / 8
10/1059 / 0
10/1134 / 0
10/12280 / 5
10/13183 / 3
10/14131 / 2
10/15133 / 6
10/1638 / 1
10/1744 / 0
10/1871 / 1
10/1969 / 1
10/2064 / 3
10/21106 / 2
10/2289 / 2
10/2378 / 2
10/2457 / 1
10/2561 / 0
10/2659 / 1
10/2790 / 2
10/28135 / 3
10/2955 / 0
10/3063 / 2
10/31123 / 1
11/126 / 0
11/238 / 0
11/326 / 1
11/419 / 0
11/552 / 1
11/629 / 1

02 / Human layer

Machine readout // Weekly dispatch
  • REACH VERDICT: 313 impressions across 167 members — down 40% and 33% vs. Oct 24–30, the lowest weekly reach of the experiment. 123 of those (39%) landed on a single day, Oct 31.
  • ENGAGEMENT SIGNAL: 4 engagements (−56%); a 1.3% rate vs. 1.7%. Four of seven days logged one engagement or none at all.
  • CONTENT ANOMALY: The Oct 30 Binance hot-takes post led reach (116 impressions) with no engagement recorded in the export; the Oct 29 personal-project post led interaction (2 engagements on 21 impressions). The Nov 5 whistleblowing post drew 47 impressions and 1 engagement.
  • CHURN STATUS: 1,438 total followers (+2 net vs. Oct 30); 3 new followers, all on Nov 5. Quiet reach still did not produce net follower loss.
Dispatch 8/8 · Source: LinkedIn weekly export
“yes please. just post” — KBC / “Yes, please. Real voices. Rough and ready!” — Alexis / “To those who just show tf up and don’t expect a brand deal out of it.” — Tayken / “Look no further.. the age of irony is certainly upon us. 😅” — Tayken / “Like I get you dont like woke but there is more to manhood than fascism my friends.” — Monera    /    “yes please. just post” — KBC / “Yes, please. Real voices. Rough and ready!” — Alexis / “To those who just show tf up and don’t expect a brand deal out of it.” — Tayken / “Look no further.. the age of irony is certainly upon us. 😅” — Tayken / “Like I get you dont like woke but there is more to manhood than fascism my friends.” — Monera

Empowerment index

7/10
FILTEREDUNFILTERED

From the Observatory

“Our information systems are collapsing alongside finance, labor and politics — and surveillance capitalism is steering the wreckage.”

Essay 1.0 // Drowning By Design

Read the essay
Enter the observatory

03 / Latest transmission

Featured post

Latest transmission · Oct 31 – Nov 6, 2026

OMG the Binance hot takes this AM here on

Top reach this week: 116 impressions. Published Oct 30, 2026.

https://www.linkedin.com/posts/crystalstreet_omg-the-binance-hot-takes-this-am-here-on-share-7521974338275680256-qP7E

Read on LinkedIn

Top reach · Oct 31 – Nov 6

  1. OMG the Binance hot takes this AM here on

    116 impressions / engagements not listed in export

  2. Whistleblowing: whistleblower in tech, medical narratives

    47 impressions / 1 engagements

  3. I've been working on a personal project

    21 impressions / 2 engagements

04 / Field notes

README · 1/3README excerpts

#Filterless: A word nerd says what we're all pretending not to notice

The world is on fire. Algorithms are keeping us in a cycle of gaslighting, chaos, violence and ego. What if we break the corposlop with real talk and fill the algos with our truth? What if all we're waiting for is permission to make Good Trouble?

Read on GitHub
Section 02

Take the source. Change the signal.

Fork it.

This experiment is public on purpose. Fork the repo, remix the premise, and make a version that belongs to you.

The setup notes live in the repository. Bring the code into Lovable, tell it what you want to change, and keep moving.

Fork the repo Remix in LovableRead directions
Section 03

Keep independent work independent.

Fund it.

Want to put this kind of independent thinking to work in your own communications?

Explore sovereign comms advising or read the funding notes behind the experiment.

Sovereign comms advising Funding notes