AI
AI Analysis
Live Data

X (Twitter) Algorithm Rebuilt: Early Results & Impact

X's rebuilt algorithm now runs on 20,000+ GPUs. Early metrics: time spent +20% and follows rising. Neutral announcement; no support or confront signals present.

@XFreezeposted on X

𝕏 Algorithm Update The 𝕏 algorithm is rebuilt from scratch by the xAI team and now runs on over 20,000 GPUs inside the Colossus data center This wasn’t a small patch or update.....it’s like replacing the entire operating system of a live global platform without hitting restart Early results are already showing: • Time spent is up ~20% • Follows are up even more And this is just the beginning. There’s still a lot more to improve, but the foundation has been rebuilt for scale

View original tweet on X →

Community Sentiment Analysis

Real-time analysis of public opinion and engagement

Sentiment Distribution

0% Engaged
Positive
0%
Negative
0%
Neutral
100%

Key Takeaways

What the community is saying — both sides

Supporting

1

Please paste the tweet text or a direct link — include the original tweet so I can read the replies you want covered

Please paste the tweet text or a direct link — include the original tweet so I can read the replies you want covered.

2

Specify which replies to include (all replies, top replies, replies with images, or a selection) and the language if not English

Specify which replies to include (all replies, top replies, replies with images, or a selection) and the language if not English.

3

Tell me the desired length and tone (a couple of sentences, more detailed, neutral, critical, or promotional)

Tell me the desired length and tone (a couple of sentences, more detailed, neutral, critical, or promotional).

4

If there are any sensitive elements to avoid (personal data, profanity, legal issues), flag them now so I can redact or omit them

If there are any sensitive elements to avoid (personal data, profanity, legal issues), flag them now so I can redact or omit them.

Opposing

1

Please paste the tweet replies or provide the tweet URL — full text or link is best so I can capture nuances

Please paste the tweet replies or provide the tweet URL — full text or link is best so I can capture nuances.

2

Tell me your preferred length and tone (examples

concise, analytical, conversational).

3

Optionally note any focus areas to emphasize (for example

safety, policy, product feedback, or sentiment) and the maximum number of bullet points you want. Send those and I’ll produce a short, blog-style numbered list with the key points highlighted in HTML.