akiboeMachine learning
AI & tech
Part of the AI family:AI agents #21All of AI →
- #1today
- #8this week
- #2this month
Interest, last 52 weeks
−1%All sources combined · vs its usual level · 100 = this series’ peak
Why it’s moving
A change counts only when two or more sources agree. “Coincided with” is timing, not proof of cause.
Doubt in searches
28% of suggestion slots for “is it worth it / dead / will AI replace it / why so hard”
Compared with: ai, deep learning, ai vs deep learning, generative ai, llm
Where it’s rising
Change vs its usual level, per source
- Bluesky+48%
- Wikipedia+3%
- Hacker News−5%
- Reddit communities−7%
- Stack Exchange−39%
Sub-skills on the move
- Computer vision−5%
- Natural language processing+32%
- Reinforcement learning+16%
- Deploying ML models+12%
- Fine-tuning models−15%
Among top-rated players
Computer-science authors with an h-index above 50 (7,897 of 11.3M, about the top 0.07%), compared with all new works with a computer-science primary topic (271,627).
- Reinforcement Learning in Robotics 2.3×
4% of them vs 2% of everyone else · 1.5×–3.4× as common (95% range) · 24 of them
h-index rewards long careers, so this cohort skews senior; a paper by several cohort authors can count more than once; only works with a computer-science primary topic are compared; topics are OpenAlex's own labels. It shows what leading researchers work on, not what anyone should learn. Common, not causal.
Papers featured on Hugging Face Daily Papers (914, about the top 6% of new arXiv AI papers), compared with all new arXiv papers in cs.AI, cs.LG, cs.CL, cs.CV, cs.RO (14,527).
- Reinforcement learning 1.6×
13% of them vs 8% of everyone else · 1.4×–1.9× as common (95% range) · 118 of them
Daily Papers are picked and upvoted by a community, so they show what researchers notice, not what is best. Topics are matched by keyword (a regex for the featured papers, arXiv's abstract search for all papers), so shares are approximate. Common, not causal.
Rising together
Shape your machine learning goal
Pick what you want to do with it. akiboe builds your path and your feed around that.
Machine learning, starting from zero, in 6 weeks.
Start learning with this goal →5 sources · high confidence · updated 7 min ago · aggregates only · How we rank