GS Paper 3
Syllabus: Science and Technology- Developments and their Applications and Effects in Everyday Life
Source: TH
Context: Google warned of the pitfalls of artificial intelligence (AI) in chatbots, as it can sometimes lead to “hallucination”.
Background: These reports emerged as OpenAI (ChatGPT), Google (Bard), and Microsoft (Bing’s beta) were opening up their AI-enabled chatbots for test users.
What are hallucinating chatbots?
- AI chatbots are trained to have human-like conversations using a process known as natural language processing (NLP).
- With NLP, they can interpret human language as it is written, which enables them to operate more or less on their own.
- Hallucination in AI chatbots is when a machine provides convincing but completely made-up answers (untrue facts). It is not a new phenomenon.
- For example, after being live on Twitter for just 24 hours in 2016, Microsoft’s chatbot Tay started parroting racist and misogynistic slurs back at users.
Why do AI chatbots start hallucinating?
- Because these models require the capability to rephrase, summarise and present intricate tracts of text without constraints.
- Facts, not being sacred, can be treated in contextual form when sifting through the information, as these models are unable to distinguish between contextual information and facts.
- For example, when asked, “What does Albert Einstein say about black holes?” AI models can return a quote made famous on the Internet rather than factual information.
- The problem becomes acute when complex grammar source material is used.
The biggest challenge: Identification of hallucinated texts
Way forward:
- A method to recognize hallucinated output and include filters in AI models to recognize and eliminate such texts is being researched.
- Tabulate and collect hallucinated texts from AI models.
Related news: India’s AI penetration factor at 3.09, highest among all G20, OECD countries
Source: TH
Context: According to a Nasscom report titled, “State of Data Science and AI Skills in India,” India ranked first in terms of AI skill penetration and AI talent concentration and 5th in AI scientific publications.
Highlights of the report:
● AI today has become the foundational technology for the world’s digital transformation.
● AI-led disruption in India’s core sectors alone has a potential impact of $500 billion on India’s economy in GVA (gross value added) terms by 2026.
● The current demand-supply gap in AI and Data Science (DS&AI) stands at 51%, significantly lower than major economies such as the US.
● As of August 2022, India’s total demand for DS&AI stood at 629K with an installed talent base of 416K.
● The country was expected to be requiring over 1 million professionals equipped with DS&AI skill sets by 2024.
Related news: Aadhar Mitra
Source: BL
Context: Aadhaar issuing body UIDAI (Unique Identification Authority of India) has launched a new AI/ML-based chatbot called Aadhaar Mitra.
Purpose: It will let individuals check their Aadhaar enrolment status, track their Aadhaar PVC card status, and information on enrolment centres.
Insta Links:
Prelims Links: (UPSC 2020)
With the present state of development, Artificial Intelligence can effectively do which of the following?
- Bring down electricity consumption in industrial units
- Create meaningful short stories and songs
- Disease diagnosis
- Text-to-speech conversion
- Wireless transmission of electrical energy
Select the correct answer using the code given below:
-
- 1, 2, 3, and 5 only
- 1, 3, and 4 only
- 2, 4, and 5 only
- 1, 2, 3, 4 and 5
Ans: 2 (UPSC Official Ans Key)
