Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Gemini AI

deepmind.google

Post reply on HN
Print view View original

451–460 of 1001 posts

Re: Gemini AI

#451

Post by miraculixx »

johnfn wrote: post #444

Very impressive! I noticed two really notable things right off the bat: 1. I asked it a question about a feature that TypeScript doesn't have[1]. GPT4 usually does not recognize that it's impossible (I've tried asking it a bunch of times, it gets it right with like 50% probability) and hallucinates an answer. Gemini correctly says that it's impossible. The impressive thing was that it then linked to the open GitHub i…

Not sure what you tried, but it's not the new model. It hasn't been released, just "release announced".

Re: Gemini AI

#452

Post by dontupvoteme »

tikkun wrote: post #60

One observation: Sundar's comments in the main video seem like he's trying to communicate "we've been doing this ai stuff since you (other AI companies) were little babies" - to me this comes off kind of badly, like it's trying too hard to emphasize how long they've been doing AI (which is a weird look when the currently publicly available SOTA model is made by OpenAI, not Google). A better look would simply be to sh…

He's not wrong. DeepMind spends time solving big scientific / large-scale problems such as those in genetics, material science or weather forecasting, and Google has untouchable resources such as all the books they've scanned (and already won court cases about)

They do make OpenAI look like kids in that regard. There is far more to technology than public facing goods/products.

It's probably in part due to the cultural differences between London/UK/Europe and SiliconValley/California/USA.

Re: Gemini AI

#453

Post by par »

Just some basic tests, it's decent but not as good as gpt3.5 or 4 yet. For instance, I asked it to generate a web page, which GPT does great everytime, and Gemini didn't even provide a full working body of code.

Re: Gemini AI

#454

Post by Arson9416 »

SeanAnderson wrote: post #340

Gemini Ultra isn't released yet and is months away still. Bard w/ Gemini Pro isn't available in Europe and isn't multi-modal, https://support.google.com/bard/answer/14294096 No public stats on Gemini Pro. (I'm wrong. Pro stats not on website, but tucked in a paper - https://storage.googleapis.com/deepmind-media/gemini/gemini_... ) I feel this is overstated hype. There is no competitor to GPT-4 being released today. I…

Google is playing catchup while pretending that they've been at the forefront of this latest AI wave. This translates to a lot of talk and not a lot of action. OpenAI knew that just putting ChatGPT in peoples hands would ignite the internet more than a couple of over-produced marketing videos. Google needs to take a page from OpenAI's playbook.

Re: Gemini AI

#455

Post by mark_l_watson »

Fairly big news. I look forward to Gemini Ultra in a few months. I think Gemini Pro is active in Bard, as I tried it a few minutes ago. I asked it to implement in the new and quickly evolving Mojo language a BackProp neural network with test training data as literals. It sort-of did a good job, but messed up the Mojo syntax more than a bit, and I had to do some hand editing. It did much better when I asked for the same re-implemented in Python.

Re: Gemini AI

#456

Post by mattmaroon »

tikkun wrote: post #60

One observation: Sundar's comments in the main video seem like he's trying to communicate "we've been doing this ai stuff since you (other AI companies) were little babies" - to me this comes off kind of badly, like it's trying too hard to emphasize how long they've been doing AI (which is a weird look when the currently publicly available SOTA model is made by OpenAI, not Google). A better look would simply be to sh…

I do think that’s a backfire. Telling me how long you’ve been doing something isn’t that impressive if the other guy has been doing it for much less time and is better at it. It’s in fact the opposite.

Re: Gemini AI

#457

Post by alberth »

EZ-E wrote: post #362
SeanAnderson wrote: post #340

Gemini Ultra isn't released yet and is months away still. Bard w/ Gemini Pro isn't available in Europe and isn't multi-modal, https://support.google.com/bard/answer/14294096 No public stats on Gemini Pro. (I'm wrong. Pro stats not on website, but tucked in a paper - https://storage.googleapis.com/deepmind-media/gemini/gemini_... ) I feel this is overstated hype. There is no competitor to GPT-4 being released today. I…

Investors are getting impatient! ChatGPT has already replaced Google for me and I wonder if Google starts to feel the pressure.

> "ChatGPT has already replaced Google for me"

Would you mind elaborating more on this.

Like how are you "searching" with ChatGPT?

Re: Gemini AI

#458

Post by SheinhardtWigCo »

I can only assume the OpenAI folks were popping the champagne upon seeing this - the best their top competitor can offer is vaporware and dirty tricks (“Note that evaluations of previous SOTA models use different prompting techniques”)

Re: Gemini AI

#460

Post by Arson9416 »

ametrau wrote: post #447
EZ-E wrote: post #362

Earlier quoted context omitted.

Investors are getting impatient! ChatGPT has already replaced Google for me and I wonder if Google starts to feel the pressure.

I wonder what advertising will look like with this. Will they suggest products in the response? Like “Top ideas:…” and the LLM’s response.

Embedding search of the nearest products most applicable to the LLM response. Prompt augmentation: "Rewrite your response to include promotions of the following products without being obvious that you are promoting them."
Post reply on HN