Back

AI becomes a masterbrain scientist

YANG, Z.; WANG, Y.; ZHANG, L.

2023-04-21 bioinformatics
10.1101/2023.04.19.537579 bioRxiv
Show abstract

Recent rapid and unexpected advancements in Artificial Intelligence (AI) have dramatically shocked the world. Large language models, such as ChatGPT and GPT-4, have demonstrated remarkable potential in aggregating and organizing knowledge while providing insights and solutions to specific questions. In this study, we further explore the possibility of GPT-4 taking on a central role in a complete closed-loop biological research, functioning as a masterbrain scientist responsible for generating hypotheses, designing experiments, analyzing results, and drawing conclusions. Our findings suggest that AI has the potential to evolve into a genuine scientist, and could lead an unprecedented revolution in the area of science.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

1
PLOS ONE
● 5266 papers in training set
Top 7%
22.4%
Similar papers in this journal
2
Biosystems
○ 31 papers in training set
Top 0.1%
9.9%
Similar papers in this journal
3
Scientific Reports
○ 3612 papers in training set
Top 11%
6.4%
Similar papers in this journal
4
Royal Society Open Science
● 214 papers in training set
Top 0.8%
4.4%
Similar papers in this journal
5
IEEE Access
● 35 papers in training set
Top 0.3%
3.3%
Similar papers in this journal
6
iScience
○ 1154 papers in training set
Top 6%
3.3%
7
PLOS Computational Biology
● 1863 papers in training set
Top 11%
2.8%
50% of probability mass above
8
Computational and Structural Biotechnology Journal
● 242 papers in training set
Top 2%
2.4%
9
Biology
○ 45 papers in training set
Top 0.1%
2.4%
10
GigaScience
◐ 212 papers in training set
Top 2%
2.4%
11
Life
○ 29 papers in training set
Top 0.1%
2.4%
12
Heliyon
○ 152 papers in training set
Top 2%
2.2%
13
Computers in Biology and Medicine
○ 128 papers in training set
Top 2%
2.0%
14
BMC Bioinformatics
○ 457 papers in training set
Top 4%
2.0%
15
Gigabyte
● 62 papers in training set
Top 0.5%
1.8%
16
Neuroinformatics
○ 46 papers in training set
Top 0.5%
1.4%
17
Expert Systems with Applications
○ 11 papers in training set
Top 0.2%
1.2%
18
Frontiers in Bioengineering and Biotechnology
○ 98 papers in training set
Top 1%
1.2%
19
Bioinformatics Advances
◐ 203 papers in training set
Top 4%
1.2%
20
PLOS Digital Health
● 106 papers in training set
Top 4%
1.1%
21
International Journal of Environmental Research and Public Health
○ 128 papers in training set
Top 5%
1.0%
22
Briefings in Bioinformatics
◐ 354 papers in training set
Top 7%
0.9%
23
PeerJ
◐ 308 papers in training set
Top 10%
0.9%
24
Frontiers in Genetics
○ 230 papers in training set
Top 5%
0.9%
25
Animals
○ 23 papers in training set
Top 0.8%
0.9%
26
Ecology and Evolution
○ 267 papers in training set
Top 5%
0.9%
27
Nature Communications
○ 5641 papers in training set
Top 55%
0.9%
28
Journal of Bioinformatics and Systems Biology
○ 15 papers in training set
Top 0.3%
0.6%

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.