Google Bard Now Better At Math & Logic By Using PaLM, Google Says

Apr 3, 2023 • 7:51 am | comments (1) by | Filed Under Google Search Engine

Google Bard Robot Cap Gowns Upgrade

Jack Krawczyk from Google, who is the lead on Bard, said on Friday that Bard just rolled out improvements around math and logic. It incorporated PaLM, Google's language model, into Bard, which helped it achieve these upgrades.

"Now Bard will better understand and respond to your prompts for multi-step word and math problems, with coding coming soon. This improvement makes Bard even more useful to people using it," Jack Krawczyk added.

Did you notice any improvements over the weekend?

Google's CEO, Sunday Pichai, was recently interviewed by the NY Times where he said these improvements were coming, I actually did a tweet thread covering the points that interested me on the topic on Friday. Here is the first tweet in the thread but I'll summarize it below:

Sundar was not surprised by the less-than-stellar reaction Google received on Bard's release because they used a toned-down LaMDA model. He was however surprised with how much of a splash ChatGPT made, he said "the reception from users, I think, was a pleasant surprise for — maybe even for them, and for a lot of us." He said it is kind of like putting a Honda Civic against a powerful race car. But it will get better fast, he said.

Sundar also didn't outright say that Bard won't be able to replace search, he said things will adapt but didn't deny it won't replace Search, even though Jack said that before.

Sundar said "Sergey [Brin, the Google co-founder] has been hanging out with our engineers for a while now." and Sergey said "it's the most exciting thing he has seen in his lifetime..."

In terms of publisher concerns and the ecosystem. Sundar Pichai said, "I think we’ll be committed to getting it right with the publisher ecosystem." Bard will evolve, "so you will see us thoughtfully evolve there as well," he added.

Yes, Google will continue to adapt Bard and try to improve it based on feedback. We see it happening and we see it happen with both ChatGPT and Bing Chat and others:

Again, curious if any of you saw an upgrade here in your tests?

Forum discussion at Twitter.

Previous story: Bing Chat Adds Images & Videos, Improves Local Results & More Changes
Ninja Banner
blog comments powered by Disqus