AI Research Answer

What is artificial intelligence?

6 cited papers · March 16, 2026 · Powered by Researchly AI

🧠

TL;DR

Artificial Intelligence (AI) is a broad field concerned with enabling machines to exhibit intelligent behaviour across a range of tasks. The current excitement…

Artificial Intelligence (AI) is a broad field concerned with enabling machines to exhibit intelligent behaviour across a range of tasks.¹The current excitement about AI, particularly machine learning (ML), has driven explosive advances in areas such as game playing, robotics, computer vision, speech recognition, and natural language processing.²

1

The Artificial Scientist: Logicist, Emergentist, and Universalist Approaches to Artificial General IntelligenceMichael Timothy Bennett, Yoshihiro Maruyama2021arXiv

2

The promise of artificial intelligence in chemical engineering: Is it here, finally?Venkat Venkatasubramanian2018AIChE Journal

Artificial General Intelligence (AGI) — A form of AI aimed at replicating broad, general cognitive abilities; approaches include logicist, emergentist, and universalist paradigms, with researchers concluding that a unified or hybrid approach is necessary.

¹²³

1

The Artificial Scientist: Logicist, Emergentist, and Universalist Approaches to Artificial General IntelligenceMichael Timothy Bennett, Yoshihiro Maruyama2021arXiv

2

Watershed of Artificial Intelligence: Human Intelligence, Machine Intelligence, and Biological IntelligenceLi Weigang, Liriam Enamoto et al.2021arXiv

3

The promise of artificial intelligence in chemical engineering: Is it here, finally?Venkat Venkatasubramanian2018AIChE Journal

Categories of AI (AHI, AMI, ABI) — AI can be divided into Artificial Human Intelligence (AHI), Artificial Machine Intelligence (AMI), and Artificial Biological Intelligence (ABI), distinguished by whether they are human-oriented, machine-oriented, or biologically-oriented, and by whether they use one/few or large samples for learning.

²¹⁴³

4

Tests of Machine IntelligenceShane Legg, Marcus Hutter2007arXiv

Tests of Machine Intelligence — Multiple tests have been proposed to define and measure machine intelligence, of which the Turing Test and its derivatives are the most widely known, though many alternatives exist.

⁴²

Deep Artificial Neural Networks (DANNs) — A popular AI approach in which multi-layer architectures, activation functions, and loss functions define the learning task; node responses are encoded as linear superpositions of neural activity, with non-linearity triggered by activation functions.

¹

Want to research your own topic? Try it free →

Diagram

 Input Data
 │
 ▼
┌─────────────────────────┐
│ Data Preprocessing │
└────────────┬────────────┘
 │
 ▼
┌─────────────────────────┐
│ Learning Algorithm │
│ (e.g., DANN layers, │
│ activation functions) │
└────────────┬────────────┘
 │
 ▼
┌─────────────────────────┐
│ Model / Knowledge Base │
└────────────┬────────────┘
 │
 ▼
┌─────────────────────────┐
│ Output / Decision │
│ (classification, │
│ detection, language) │
└─────────────────────────┘

Table

Category	Orientation	Sample Requirement	Example Application
AHI	Human-oriented	One/few-shot	Image classification
AMI	Machine-oriented	Large samples	Robotics, NLP
ABI	Biologically-oriented	Varies	Biological modelling

AI systems such as AlphaGo, autonomous cars, and natural language processors represent stunning recent advances, though there is also considerable hype and a tendency to overestimate their promise.¹The Turing Test specifically checks for human-level intelligence through repeated interaction requiring learning and adaptation to a conversation partner, rather than testing any putative general intelligence.²³

1

The promise of artificial intelligence in chemical engineering: Is it here, finally?Venkat Venkatasubramanian2018AIChE Journal

2

Learning, Social Intelligence and the Turing Test - why an "out-of-the-box" Turing Machine will not pass the Turing TestBruce Edmonds, Carlos Gershenson2012arXiv

3

The Artificial Scientist: Logicist, Emergentist, and Universalist Approaches to Artificial General IntelligenceMichael Timothy Bennett, Yoshihiro Maruyama2021arXiv

Want to research your own topic? Try it free →

Despite impressive advances, AI research has historically been subject to cycles of hype and overestimation, as observed with expert systems in the 1980s and neural networks in the 1990s.

¹²

The Turing Test and most machine intelligence tests are limited in scope; few researchers are even aware of alternatives beyond the Turing Test and its derivatives, indicating a gap in how machine intelligence is evaluated.

1

The promise of artificial intelligence in chemical engineering: Is it here, finally?Venkat Venkatasubramanian2018AIChE Journal

2

The Deep Arbitrary Polynomial Chaos Neural Network or how Deep Artificial Neural Networks could benefit from Data-Driven Homogeneous Chaos TheorySergey Oladyshkin, Timothy Praditia et al.2023arXiv

AI encompasses a wide spectrum of approaches, from narrow machine learning to ambitious AGI research requiring hybrid methodologies.

¹²

AI can be meaningfully categorised into human-oriented, machine-oriented, and biologically-oriented branches, each with distinct learning paradigms.

³²

Deep neural networks are among the most popular current AI tools, relying on layered architectures and activation functions for non-linear learning.

⁴

Measuring machine intelligence remains an open challenge, with the Turing Test being only one of many proposed evaluation frameworks.

⁵³²

Real-world AI applications span game playing, computer vision, speech recognition, and NLP, but hype continues to outpace practical capabilities in many domains.

1

The Artificial Scientist: Logicist, Emergentist, and Universalist Approaches to Artificial General IntelligenceMichael Timothy Bennett, Yoshihiro Maruyama2021arXiv

2

Learning, Social Intelligence and the Turing Test - why an "out-of-the-box" Turing Machine will not pass the Turing TestBruce Edmonds, Carlos Gershenson2012arXiv

3

Watershed of Artificial Intelligence: Human Intelligence, Machine Intelligence, and Biological IntelligenceLi Weigang, Liriam Enamoto et al.2021arXiv

4

The Deep Arbitrary Polynomial Chaos Neural Network or how Deep Artificial Neural Networks could benefit from Data-Driven Homogeneous Chaos TheorySergey Oladyshkin, Timothy Praditia et al.2023arXiv

5

Tests of Machine IntelligenceShane Legg, Marcus Hutter2007arXiv

Want to research your own topic? Try it free →

"Comparison of Artificial General Intelligence approaches: logicist vs emergentist vs connectionist"
"One-shot and few-shot learning methods in deep neural networks"
"Evaluation benchmarks for machine intelligence beyond the Turing Test"

Research smarter with AI-powered citations

Researchly finds and cites academic papers for any research topic in seconds. Used by students across India.

Try Free See Pricing