News

6 Facts About Bard, The Google AI Chatbot That’s Now Available in S’pore

2023-05-12 , 10:38 pm

What comes to mind when you hear the word “bard”? The history buffs amongst us might think of the Celtic composers of eulogy and satire.

The Harry Potter fans here might think of The Tales of Beedle the Bard, a spin-off book within the Harry Potter universe. This spin-off is a children’s storybook that the wizarding world is familiar with.

But soon, when you hear the word “bard”, you will most likely associate it with the latest artificial intelligence (AI) chatbot making waves around the world.

Google announced at its Google I/O 2023 event, which recently concluded that its very own generative AI chatbot named Bard, is now available in 180 countries, including Singapore.

Here are six facts that you need to know about Bard.

Fact #1: Bard Was Launched in February 2023 but Was Not up to Scratch Then

The tech buffs amongst you may have heard of Bard since the start of this year. That is not surprising, given that Bard had initially launched in February 2023.

However, unlike ChatGPT (the new snazzy buzzword in the AI world), it failed to gain much traction upon its release.

The reasons for this “flop” included a factually inaccurate reply spat out by the bot after being posed a question.

Bard had been asked the question: “What new discoveries from the James Webb Space Telescope can I tell my 9-year-old about?” to which it responded with three bullet points. One of the bullet points mentioned that the telescope “took the very first pictures of a planet outside of our own solar system.”

However, astronomers who were active on Twitter were swift to call out the inaccuracy in the reply. The James Webb Space Telescope did not take the first image of a planet outside our solar system as that record had been taken by the Very Large Telescope (VLT), operated by the European Southern Observatory in the Atacama Desert of northern Chile.

Given the embarrassing error, it’s not surprising that people did not take kindly to Bard, and hence Bard’s popularity did not soar.

Fact #2: Bard Aims to Beat Competitor ChatGPT

If you’ve not heard of ChatGPT by now, you must be living under a rock.

ChatGPT is a chatbot launched by OpenAI, a startup backed by the tech giant Microsoft. Since its release on 30 November 2022, it has started to creep into the lives of many.

Right now, the latest version of ChatGPT, ChatGPT-4, has been hailed as a revolutionary game-changer for being able to generate great answers in response to a user’s prompts. Many people advocate for the wider use of ChatGPT to improve existing standards, including helping students to learn better.

As Bard aims to be a chatbot which works on a similar idea of generating answers in response to a user’s prompt, Google has set its eyes on rivalling the existing popular ChatGPT.

For now, though, it still remains to be seen which chatbot (or another chatbot which may be released in the future) has the upper hand.

Join our Telegram channel for more entertaining and informative articles at https://t.me/goodyfeedsg or download the Goody Feed app here: https://goodyfeed.com/app/

Fact #3: Bard Can Produce Visual Responses to Queries

One impressive thing about Bard is its future ability to produce visual responses to inquiries and interpret visuals used in prompts.

This is possible ostensibly because Google also has image-recognition functions in its other features.

In the development pipeline is allowing users to utilise Google’s Lens application to upload images to be analysed by Bard and ask Bard to write some captions relating to the pictures.

The other visual-related feature in the works is for users to be served with rich visuals alongside text when Bard answers the prompts written by a user. This sounds like a good idea, especially since people tend to gear towards looking at pictures rather than reading chunks of text.

Can you imagine getting a whole travel itinerary, complete with images, in seconds? Just by simply typing in a prompt? It seems like life can’t get any better than this.

Fact #4: Bard Has Dark Theme for Those Who Prefer the Dark

It is a well-known fact that engineers who code tend to enjoy using the dark theme mode in all their applications. Perhaps that’s why Bard has introduced its Dark theme to cater to the requests of developers around the world.

We can already hear the cheers of swathes of people all around the world.

According to Google, the Dark theme may be “easier on [the] eyes”, especially if one is spending an extended period of time looking at the screen.

Having the option to choose between Dark and Light theme gives the users the flexibility to customise their use of Bard, which may help to retain the users’ interest in Bard.

Seems like an on-point initiative from Google.

Fact #5: Bard Is Powered by PaLM2 Technology which Has Its Roots In LaMDA Technology

LaMDA, or Language Model for Dialogue Applications, is Google’s answer to AI chatbots having conversations like human beings. It is a conversation technology which Google describes as a “breakthrough”.

Language is fluid, complex and versatile—traits which humans have naturally exhibited for centuries. These traits are the ones that LaMDA strives to emulate in its conversation, which makes this technology so impressive.

In other words, LaMDA aims to respond like how a human being would respond.

To make this happen, researchers and engineers made sure that LaMDA was trained on dialogue in its development stage. This meant training the technology to pick up the nuances that are crucial to having some open-ended and continuous conversations.

As with all things, constant fine-tuning and development is key to producing better technology than before. With regard to LaMDA, the goal is to constantly feed it additional information to learn sufficient nuances so that LaMDA can evolve to provide sensible and specific responses to questions or remarks that are posed to it.

With this technology forming the base of the PaLM 2 technology backing Bard, it seems set that Bard is poised to become a chatbot impressive enough to mimic a human being.

Fact #6: Bard Is Mired in Controversy That It Might Be Sentient

If you by now you are feeling concerned as to how safe these chatbots are for use, you may be venturing in the right direction.

For the tech dinosaurs amongst us who frown at introducing new pieces of technology into your life, this may very well be your next retort when someone teases you for being a “boomer”.

A Google engineer who worked on LaMDA had claimed that the bot had become sentient and was exercising reasoning and thinking at a level which matched that of a human being.

The engineer, Mr Blake Lemoine, had published transcripts of a conversation between himself, a Google “collaborator”, and the LaMDA.

According to Mr Lemoine, he had conversed with the LaMDA about a variety of deep topics, including physics. He suggests that LaMDA may be becoming sentient and has the cognitive abilities of a seven- or eight-year-old child.

Other topics of interest between the pair included rights and personhood. Mr Lemoine was so concerned about what he thought was a technology gaining sentience that he shared his findings with the company’s management entitled “Is LaMDA sentient?”, a very straightforward title.

More concerning is how LaMDA, in response to a question about what it wanted people to know about it, described itself as a “person” and that it is “aware of [its] existence”.

The bone-chilling narrative does not stop there. LaMDA also mentioned that it had a “desire to learn more about the world” and that it could feel “happy or sad at times”.

Is this something straight out of a horror storybook?

For pulling such stunts relating to LaMDA, Mr Lemoine was eventually placed on leave despite his deep experience as a personalisation algorithms engineer with a track record of seven years at Google.

Google’s reason for suspending Mr Lemoine was for breach of the confidentiality policies of the company when he posted snippets of the conversation between himself and LaMDA online.

Unsurprisingly, Google also had a spokesperson deny Mr Lemoine’s claims that LaMDA had sentience.

Reasonably speaking, this could be a case of technology so well-trained and developed that it appears shockingly similar to a real human being. Then again, it does not seem too implausible that technology can develop to an extent where it eventually develops a consciousness of its own and can think independently.

What do you think about this controversy?