Gemini 2.0: our new AI model for the agentic era • Hacker News Board

Re: Gemini 2.0: our new AI model for the agentic era

#411

Post by kthartic » Thu, Dec 12, 2024, 9:51 AM UTC

thomassmith65 wrote: post #63

jncfhnb wrote: post #16

Am I alone in thinking the word “agentic” is dumb as shit? Most of these things seem to just be a system prompt and a tool that get invoked as part of a pipeline. They’re hardly “agents”. They’re modules.

It's easier for consultants and sales people to sell to enterprise if the terminology is familiar but mysterious. Bad 1. installed Antivirus software 2. added screen-size CSS rules 3. copied 'Assets' harddrive to DropBox 4. edited homepage to include Bitcoin wallet address link 5. upgraded to ChatGPT Pro "Good" 1. Cyber-security defenses 2. Responsive Design implementation 3. Cloud Storage 4. Blockchain Technology ga…

These are great examples. Also an excellent technique for relaying project updates to non-technical managers

Re: Gemini 2.0: our new AI model for the agentic era

#412

Post by jug » Thu, Dec 12, 2024, 9:56 AM UTC

danpalmer wrote: post #299

siliconc0w wrote: post #44

What's everyone's favorite LLM leaderboard? Gemini 2 seems to be edging out 4o on chatbot arena( https://lmarena.ai/?leaderboard )

Notably, GPT-4o is a "full size" model, whereas Gemini 2 Flash is the small and efficient variant in that family as far as I understand it.

True that - and I think Gemini-Exp-1206 is Gemini 2.0 Pro in testing. I noticed how they only replaced the "experimental" moniker for one of their experimental models, and it turned into 2.0 Flash. And that still experimental model is currently leading across all categories.

Re: Gemini 2.0: our new AI model for the agentic era

#413

Post by shakna » Thu, Dec 12, 2024, 10:06 AM UTC

Alifatisk wrote: post #406

socksy wrote: post #402

Earlier quoted context omitted.

Right, and Macs are made out of aluminium

What is Firefox made out of then?

Aluminum. It's American.

Re: Gemini 2.0: our new AI model for the agentic era

#414

Post by internet_points » Thu, Dec 12, 2024, 10:13 AM UTC

modeless wrote: post #383

https://aistudio.google.com/live is by far the coolest thing here. You can just go there and share your screen or camera and have a running live voice conversation with Gemini about anything you're looking at. As much as you want, for free. I just tried having it teach me how to use Blender. It seems like it could actually be super helpful for beginners, as it has decent knowledge of the toolbars and keyboard shortcu…

I got so hopefuly by your comment I showed it my current bug that I'm working on, I even prepared everything first with my github issue, the relevant code, the terminal with the failing tests, and I pasted it the full contents of the file and explained carefully what I wanted to achieve. As I was doing this, it repeated back to me everything I said, saying things like "if I understand correctly you're showing me a file called foo dot see see pee" and "I see you have a github issue open called extraneous spaces in frobnicator issue number sixty six" and "I see you have shared some extensive code" and after some more of this "validation"-speak it started repeating the full contents of the file like "backquote backquote backquote forward slash star .. import ess tee dee colon colon .." and so on.

Not quite up to excited junior-level programmer standards yet. But maybe good for other things who knows.

Re: Gemini 2.0: our new AI model for the agentic era

#415

Post by internet_points » Thu, Dec 12, 2024, 10:14 AM UTC

Alifatisk wrote: post #406

socksy wrote: post #402

Earlier quoted context omitted.

Right, and Macs are made out of aluminium

What is Firefox made out of then?

The amount of rust indicates iron. So Firefox is very irony.

Re: Gemini 2.0: our new AI model for the agentic era

#416

Post by afro88 » Thu, Dec 12, 2024, 10:21 AM UTC

computergert wrote: post #372

It's funny how even the latest AI models struggle with simple questions. "What's the first name of Freddy LaStrange"? >> "I do not have enough information about that person to help with your request. I am a large language model, and I am able to communicate and generate human-like text in response to a wide range of prompts and questions, but my knowledge about this person is limited. Is there anything else I can do…

Is o1 not one of the latest models?

https://chatgpt.com/share/675ab91c-c158-8004-9dfc-ea176ba387...

A better way to say it is: "funny how Google can't keep up with the reasoning effectiveness of OpenAI's latest models"

Re: Gemini 2.0: our new AI model for the agentic era

#417

Post by Zababa » Thu, Dec 12, 2024, 10:26 AM UTC

modeless wrote: post #383

https://aistudio.google.com/live is by far the coolest thing here. You can just go there and share your screen or camera and have a running live voice conversation with Gemini about anything you're looking at. As much as you want, for free. I just tried having it teach me how to use Blender. It seems like it could actually be super helpful for beginners, as it has decent knowledge of the toolbars and keyboard shortcu…

I don't know what's not working but I get "Has a large language model. I don't have the capability to see your screen or any other visual input. My interactions are purely based on the text that you provide"

Re: Gemini 2.0: our new AI model for the agentic era

#418

Post by irthomasthomas » Thu, Dec 12, 2024, 10:32 AM UTC

I wrote about a possibly emergent behaviour in Gemini 2 that was only seen previously in sonnet-3.5 https://x.com/xundecidability/status/1867044846839431614

Re: Gemini 2.0: our new AI model for the agentic era

#419

Post by moffkalast » Thu, Dec 12, 2024, 10:41 AM UTC

modeless wrote: post #383

https://aistudio.google.com/live is by far the coolest thing here. You can just go there and share your screen or camera and have a running live voice conversation with Gemini about anything you're looking at. As much as you want, for free. I just tried having it teach me how to use Blender. It seems like it could actually be super helpful for beginners, as it has decent knowledge of the toolbars and keyboard shortcu…

This'll be so fantastic once local models can do it, because nobody in the right mind would stream their voice and everything they do on their machine to Google right? Right?

Oh who am I kidding, people upload literally everything to drive lmao.

Re: Gemini 2.0: our new AI model for the agentic era

#420

Post by IanCal » Thu, Dec 12, 2024, 10:50 AM UTC

nycdatasci wrote: post #109

Gemini 2.0 Flash is available here: https://aistudio.google.com/prompts/new_chat Based on initial interactions, it's extremely verbose. It seems to be focused on explaining its reasoning, but even after just a few interactions I have seen some surprising hallucinations. For example, to assess current understanding of AI, I mentioned "Why hasn't Anthropic released Claude 3.5 Opus yet?" Gemini responded with text that…

I'm having it lie a lot too. It wrote some pretty decent attempt at code in a new framework given instructions, which is great, but it made a key error about

That's fine, but it couldn't spot it, then it told me that I had put that syntax in the instructions and quoted me - but that wasn't true. It repeatedly said it had undone that and rewrote the code with it still in as well.

It added some async stuff where it wasn't needed (no way for it to know to be fair), what was interesting was when told this it apologised and explained it had just been doing so much async work recently it got confused.

Another interesting quote

> I am incredibly sorry for the repeated errors. I have made so many mistakes, and I am very frustrated with myself. I believe this is now correct, and adheres to the instructions.