Google Debuts an Updated Gemini 2.5 Pro AI Model Ahead of I/O

An anonymous reader quotes a report from TechCrunch: Google on Tuesday announced the launch of Gemini 2.5 Pro Preview (I/O edition), an updated version of its flagship Gemini 2.5 Pro AI model that the company claims tops a number of widely used benchmarks. Gemini 2.5 Pro Preview (I/O edition) is available via the Gemini API and Google's Vertex AI and AI Studio platforms, and is priced the same as the Gemini 2.5 Pro model it effectively replaces. It's also in Google's Gemini chatbot app for the web and for mobile devices.

The model's release comes ahead of Google's annual I/O developer conference (hence the "I/O edition" designation), where Google is expected to unveil a host of models, as well as AI-powered tools and platforms. [...] According to Google, Gemini 2.5 Pro Preview (I/O edition) has "significantly" improved capabilities for coding and building interactive web apps. The model is also better at tasks like code transformation -- that is, modifying a piece of code to achieve a specific goal -- and code editing, the company says. Google says the Gemini 2.5 Pro Preview (I/O edition) leads the WebDev Arena Leaderboard, a benchmark measuring a model's ability to create aesthetically pleasing and functional web apps. It also achieved a score of 84.8% on VideoMME, a popular benchmark designed to evaluate the video analysis capabilities of multi-modal large language models.

Will it suck my dick?

[TheMiddleRoad]

I'm just trying to find a job that will be safe from AI.

Re:

[ebunga]

No, it will not, because that would be something at least one person wants, and Google is incapable of creating anything people actually want, or if they somehow manage to pull it off, they'll either use it to put ads somewhere they don't belong, or will cancel the product just as soon as you're really starting to like it.

Re:

[nightflameauto]

I'm just trying to find a job that will be safe from AI.

There are already robotic implements purporting to use AI that do that. Seriously. Though I doubt Google had much to do with it, since, as ebunga noted, that would be something people might actually want.

Re:

[h33t l4x0r]

Talk about a job nobody wants.

Re:

[hcs_$reboot]

It will, via a robot, just give it some more time.

putting a value on intellect

[Big Hairy Gorilla]

It sounds a lot like salary bands. You have to justify the value somehow, so, I read thru the article and deeply down there somewhere a click or tow is the leaderboard itself. Well, it's a semblance of a blind measurement. I'm pretty sure it's worth something for the robots to compete, but Gorillas are cynical, we know the games that go on. How do you monetize AI? use a metaphor people relate to: employees. Lay off junior. Get AI JR for hmmm... what is the right price? I guess a chatbot is the unit of pur

Gemini 2.5 Pro

[hcs_$reboot]

Used it for a while, and it's really impressive. However, maintaining a lengthy discussion, despite its 1,048,576 input tokens, makes it go into a kind of loop, or losing context.

Re:

[Big Hairy Gorilla]

is there a way to know when the AI is losing the thread of conversation before it drops the ball?
can you monitor the context window?

Re:

[SpzToid]

is there a way to know when the AI is losing the thread of conversation before it drops the ball?
can you monitor the context window?

I find it to be practical to ask the AI to summarize where we are and what have we been trying to do before 'we run out of tokens', so I can copy/paste the result to start a new thread to pick up where we left off.