AI HALLUCINATIONS
Banking on GenAI: The artificially intelligent future of finance
Top Indian banks like SBI, HDFC, and Axis are adopting GenAI for better services. Data shows that the BFSI sector's IT spending was a whopping $13.2 billion in 2024. Similarly, fintech startups like Setu, AdvaRisk, Velocity, and Gnani.ai are assisting banks with AI-led innovations like fraud detection and data analysis.
Why an ‘AI health coach’ won't solve the world's chronic disease problems
Health is an appealing direction for an AI industry that has promised to transform civilisation, but whose huge growth of the past couple of years is beginning to look like it's stalling. Companies and investors have pumped billions into the technology, but it is still often a solution looking for problems.
Flipkart in Q-fest; AI upskilling
Happy Monday! Flipkart has upped the quick commerce ante ahead of crucial festive season sale. This and more in today’s ETtech Morning Dispatch.
Job market’s new mantra: Leverage AI, stay relevant
AI courses in India see explosive demand as professionals race to upskill for lucrative roles. Companies value AI for automation and innovation, leading to salary hikes of up to 200%. Cybersecurity demand surges amidst alarming cyber threats. Upskilling platforms offer tailored AI programmes to meet the digital economy's demands. Are you ready for the AI revolution?
OpenAI working on new reasoning technology under code name 'Strawberry'
The project, details of which have not been previously reported, comes as the Microsoft-backed startup races to show that the types of models it offers are capable of delivering advanced reasoning capabilities.
Sam Altman, Arianna Huffington launch personalised AI health coach
Sam Altman, OpenAI CEO, and Arianna Huffington, founder of behaviour tech company Thrive Global, have unveiled a startup Thrive AI Health, a hyperpersonalised health coach, driven by AI. It seeks to model healthy behaviour in users based on their health data, with a special focus on chronic conditions like diabetes.
- Go To Page 1
Edtech consolidation; GenAI chatbot hallucinations
Edtech firm Schoolnet has acquired learning app Genius Teacher for an undisclosed sum. This and more on today’s ETtech Top 5.
Ghosts in the machine: Peril of hallucinations in GenAI chatbots
Hallucinations are a big issue while deploying GenAI-based chatbots. Hallucination is a phenomenon where an AI chatbot offers factually incorrect information in response to a question. How are firms navigating its pitfalls?
ET Explainer: Cloudflare's new tool aims to block AI bots from scraping website content
Cloudflare has introduced a new tool to block AI bots from scraping website content. The tool aims to protect content publishers from unauthorized use of their works to train AI models. Find out more about how this tool works and why it's important.
OpenAI releases new GPT-4 model to find errors in ChatGPT's responses
GPT-4 will help align with AI systems, through reinforcement learning from human feedback (RLHF), which helps make responses from large language models more accurate. OpenAI said that with ChatGPT becoming more accurate and errors more subtle, a model like CriticGPT became necessary to find inaccuracies.
The AI boom has an unlikely early winner: Wonky consultants
As the tech industry explores avenues to monetise generative AI, consultants have already started capitalising on its potential profitability. Companies like McKinsey & Co, IBM, and Accenture are experiencing increasing sales and expanding their workforce. This trend is driven by a desperate need for technology experts who can serve as guides, helping businesses understand the implications of generative AI and its potential benefits.
Central banks must prepare for profound impact of AI, BIS says
In its first major report about the rapidly advancing world of AI Bank for International Settlements (BIS)has said, policymakers need to harness its immense power to monitor data in real time to "sharpen" their inflation-predicting abilities. A former Swedish central banker said, "We like to hold humans accountable."
Meta AI assistant rolls out with text, and image features
Meta has launched Meta AI across WhatsApp, Facebook, and Instagram, offering text and image generation capabilities. Ryan Cairns, VP of Engineering at Meta, highlighted the advantages of Meta AI over competitors like Gemini, ChatGPT, and Copilot.
Amazon mulls $5 to $10 monthly price tag for unprofitable Alexa service, AI revamp
Known internally as 'Banyan,' the project would represent the first major overhaul of the voice assistant since it was introduced in 2014 along with the Echo line of speakers. Amazon is keeping pace with competitors in generative AI, given that Google, Microsoft, and OpenAI have received significant praise for their advanced chatbots. These systems can swiftly generate coherent responses to complex prompts or inquiries.
AI could help spread false and misleading information on Holocaust, UNESCO report warns
A UNESCO report state, artificial intelligence could potentially propagate false and misleading information about the Holocaust online. This risk arises either from flaws in AI systems themselves or from deliberate efforts by hate groups and Holocaust deniers to exploit AI to generate content that undermines historical truths about the Nazi-perpetrated murder of Jews and other targeted groups.
Eye on AI: Tech giants form industry group, Apple-OpenAI deal and other top developments
Google, Microsoft, Intel and other tech companies have come together to take on chipmaking giant Nvidia. Apple partners with OpenAI. The ChatGPT maker is training its new flagship model to succeed GPT-4. Google is putting limits on AI search experience. French unicorn startup Mistral releases generative AI model for coding.
US intelligence agencies' embrace of generative AI is at once wary and urgent
A Silicon Valley firm used generative AI in the Sable Spear operation to analyze Chinese fentanyl trafficking data, surpassing human analysis. It predicted Russia's invasion of Ukraine and alerted government customers to North Korean missile launches. U.S. intelligence agencies are exploring AI for geospatial intel and assessing AI missions.
Cats on the moon? Google's AI tool is producing misleading responses that have experts worried
Google's AI-powered search engine now provides instant answers, but errors like fake facts and debunked theories are causing concern. Experts warn of bias, misinformation, and potential dangers, especially in emergency situations.
Govt saying aye to laying strong AI foundation
The government may rope in higher education institutes, key researchers for the model that can be customised for wide range of applications. It will be trained on publicly available data, digitised records of books, journals and research papers from public libraries.
EU warns Microsoft to give Bing AI risk data or face fines
The EU has identified generative AI as one of the key threats to elections being held across the 27-country bloc June 6-9. Brussels in March pressed Bing, alongside several big digital platforms including Google Search, TikTok and YouTube, about action they have taken to mitigate AI risks.
Apple will revamp Siri to catch up to its chatbot competitors
Introduced in 2011 as the original virtual assistant in every iPhone, Siri had been limited for years to individual requests and had never been able to follow a conversation. It often misunderstood questions.
Indian gig workers toil at frontlines of AI revolution
Akash and Ikshita engage in online work. Gig workers in India train AI models through microtasks. India emerges as a data annotation hub with a potential $8.22 billion global market and a million-strong workforce by 2028.
US newspapers sue OpenAI for copyright infringement over AI training
New York Daily News, Chicago Tribune, MediaNews Group sue Microsoft, OpenAI for misusing reporters' work. Lawsuit includes Copilot, alleges copying of articles. Ongoing suits by other news outlets. Lawsuits aim for damages and infringement block.
Dictionary.com chooses 'hallucinate' as 2023's Word of the Year. Here's why
"Hallucinate" has been selected as the 2023 Word of the Year by Dictionary.com. In the context of machine learning programs, the second definition characterizes it as generating false information contrary to the user's intent and presenting it as true and factual. Grant Barrett, the head of lexicography at Dictionary.com, noted that the word's trajectory in the technology domain mirrors that of other terms like "spam" and "virus."
Chatbots may 'hallucinate' more often than many realise
When Google introduced a similar chatbot several weeks later, it spewed nonsense about the James Webb telescope. The next day, Microsoft's new Bing chatbot offered up all sorts of bogus information about the Gap, Mexican nightlife and singer Billie Eilish. Then, in March, ChatGPT cited a half dozen fake court cases while writing a 10-page legal brief that a lawyer submitted to a federal judge in Manhattan.
Mamaearth parent's muted market debut, early investors clock gains; Xpressbees bags $80 million
Honasa Consumer had a rather flat debut on the stock markets today, opening at a 2% premium to its issue price. This and more in today’s ETtech Top 5.
Businesses testing ways to safeguard against AI hallucination
Telecom, medicine and ad-tech are among sectors facing the threat of AI hallucinations that can cause bias to set into the model, making the decision-making process flawed.
Chatbots sometimes make things up. Is AI's hallucination problem fixable?
"I don't think that there's any model today that doesn't suffer from some hallucination," said Daniela Amodei, co-founder and president of Anthropic, maker of the chatbot Claude 2.
Get ready to deal with AI's hallucination problem as chatbots become part of your daily lives
Described as hallucination, confabulation or just plain making things up, it's now a problem for every business, organisation and high school student trying to get a generative AI system to compose documents and get work done.
Load More
