Future of AI: ChatGPT vs Bard battle will point to what lies ahead | Latest News India - Hindustan Times
close_game
close_game

Future of AI: ChatGPT vs Bard battle will point to what lies ahead

By, New Delhi
Feb 08, 2023 01:36 PM IST

To quickly recap, ChatGPT can hold reasonable conversations over topics, explain and contextualise matters across a wide variety of fields and churn out fiction.

This week, Google responded to the conversational AI battles by unveiling Bard, which will rival ChatGPT, the artificially intelligent text generation system by tech startup OpenAI that has become, and often generated, the talk of the town.

Artificial intelligence.(Thinkstock) PREMIUM
Artificial intelligence.(Thinkstock)

To quickly recap, ChatGPT can hold reasonable conversations over topics, explain and contextualise matters across a wide variety of fields and churn out fiction. In doing so, it serves up as a new source for information that has existed on the web, directly challenging what Google is best known for.

Unsurprisingly, Google and other tech companies have hastened the launch of such services. The global chatbot market size is expected to be worth as much as $3.99 billion by the year 2030, according to Grand View Research’s latest data. Google will have the advantage of being able to use an existing core — the Language Model for Dialogue Applications (LaMDA) platform. Bard is now available to a select group of “trusted testers” with the plan to make it widely available to the public, in the coming weeks.

The tech giant insists that they will take things slowly, if need be, which ties in with the AI Principles that Google set out in 2018, becoming one of the first tech companies to do so. “It’s critical that we bring experiences rooted in these models to the world in a bold and responsible way. That’s why we’re committed to developing AI responsibly,” said Sundar Pichai, CEO of Google and Alphabet, in a statement, unveiling Bard.

The fear of unoptimised or untested AI responding to user queries, raises genuine concern for tech companies. Chatbots have left no stone unturned either, to fuel that fear. At the heart of this is simply the concern that these tools could be considered as authentic sources of information while serving up inaccuracies that can, at its most innocuous, confuse people, but at worst, become conduits for misinformation.

The many concerns

Researchers from the Royal Holloway University of London say chatbots go beyond learning natural language and the art of conversation.

“It holds anthropomorphic features that could trigger feeling of closeness and connectedness,” says researcher Lorentsa Gkinko. “Their use could trigger a range of emotions beyond what users experience with traditional enterprise systems,” adds researcher Amany Elbanna.

It is difficult to point to the exact contours of Bard, as it will eventually emerge for public use. At best, it may be as dynamic and conversational as ChatGPT.

“ChatGPT is shockingly good at sounding convincing on any conceivable topic. But OpenAI is clear that there is no source of truth during training,” says Arvind Narayanan, professor of computer science at Princeton University. “That means that using ChatGPT in its current form would be a bad idea for applications like education or answering health questions,” he adds.

OpenAI has now announced the ChatGPT Plus subscription service, which for $20 a month (around 1,650), gets users faster responses and first access to new features as well as platform improvements.

It is important to note that Google, for now, has not mentioned a possible Bard integration within Search, something its rival Microsoft announced it will do with ChatGPT. Microsoft, already an investor in OpenAI, has plans to pump in $10 billion more and has already embarked on integrating the AI tool into existing products such as Teams.

“Modern tools powered by AI hold the promise to boost individual, team, and organizational-level productivity and fundamentally change how we work,” said Nicole Herskowitz, Vice president for Microsoft Teams, when the feature, that allows Teams to automatically take notes and create discussed tasks from a meeting, was launched.

Scope of AI chatbots

The version of Bard, going into testing now, will likely not be the same as what everyone gets to use in a few weeks’ time. For now, Google will be using what they call a lightweight version of LaMDA. It uses less computing power, which allows for more users on board and therefore more feedback on responses. That’ll change, in due course.

“Bard can be an outlet for creativity, and a launchpad for curiosity, helping you to explain new discoveries from NASA’s James Webb Space Telescope to a 9-year-old, or learn more about the best strikers in football right now, and then get drills to build your skills,” said Pichai.

That should give us a fair hint that Bard will do everything ChatGPT does, and then some, perhaps.

What will need to be closely watched is how Google expands Bard: Will it be a separate app or web-only? Will some of Google’s services such as Meet virtual meeting platform or Docs productivity suite get it?

Google’s insights into changing search trends may give Bard an advantage that ChatGPT and indeed other conversation AI platforms do not have.

Is AI ready to replace traditional search?

“When people think of Google, they often think of turning to us for quick factual answers, like “how many keys does a piano have?” But increasingly, people are turning to Google for deeper insights and understanding — like, “is the piano or guitar easier to learn, and how much practice does each need?” Pichai said, capturing the difference in paradigm that AI tools can bring to search, than conventional systems that work on scraping the web to just identify what word appears where.

But an experimental conversational AI service, much like any other AI, will only be as good as the data sets it learns from. There are often challenges with inclusivity, diversity, and sensitivity in AI conversations. Image processing algorithms have often struggled with skin tones, for instance.

This means the challenge, when a service attempts to not just find what is on the internet but understand it, is much greater and has far deeper implications.

ChatGPT in its various evolutions over time, and indeed Google’s LaMDA, have run up against these challenges. Meta’s Galactica is perhaps the shining example of a large language model, not primed for accuracy. In November last year, three days after the public release of the chatbot, the beta version was taken offline. The problem? It couldn’t separate fact from fiction, and responses were filled with misinformation and bias.

“I asked Galactica about some things I know about and I’m troubled. In all cases, it was wrong or biased but sounded right and authoritative. I think it’s dangerous,” wrote Michael Black, director, Max Planck Institute for Intelligent Systems.

Meta had more hope from the other chatbot, BlenderBot, with the company insisting multiple layers of checks and fail-safe are in place. It is only available to users in US, for now.

Microsoft itself had an unsuccessful tryst with an AI chatbot in 2016. Tay, as it was called, bolted on to Twitter (it used the @TayandYou handle) in the hope it’ll learn from humans. Soon, the bot was tweeting racist, antisemitic, and abusive responses.

Microsoft had to take it offline, a day and some 96,000 AI generated tweets later.

Google’s own try with AI in conversations, with the comparatively much simpler Allo messaging app, didn’t go according to plan either. Allo, expected to be a WhatsApp competitor, was shut down subsequently.

But much water has flown under the bridge since then, as ChatGPT has shown, and the AI arms race is expected to offer us an even more dramatic landscape in some years.

Unveiling Elections 2024: The Big Picture', a fresh segment in HT's talk show 'The Interview with Kumkum Chadha', where leaders across the political spectrum discuss the upcoming general elections. Watch now!

Get Current Updates on India News, Election 2024, Arvind Kejriwal News Live, Bihar Board 10th Result 2024 Live along with Latest News and Top Headlines from India and around the world.

Continue reading with HT Premium Subscription

Daily E Paper I Premium Articles I Brunch E Magazine I Daily Infographics
freemium
SHARE THIS ARTICLE ON
Share this article
  • ABOUT THE AUTHOR
    author-default-90x90

    Vishal Mathur is Technology Editor for Hindustan Times. When not making sense of technology, he often searches for an elusive analog space in a digital world.

SHARE
Story Saved
Live Score
OPEN APP
Saved Articles
Following
My Reads
Sign out
New Delhi 0C
Thursday, March 28, 2024
Start 14 Days Free Trial Subscribe Now
Follow Us On