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

Nano Banana 2: Google's latest AI image generation model

blog.google

Post reply on HN
Print view View original

541–550 of 593 posts

Re: Nano Banana 2: Google's latest AI image generation model

#541

Post by shinycode »

krisoft wrote: post #539
shinycode wrote: post #534

Earlier quoted context omitted.

The same kind of proof we accept for any scientific claim: converging, reproducible evidence that rules out competing explanations. Concretely, that means: We already have indirect evidence: conscious states vary predictably with brain states. Damage specific regions, lose specific functions. Alter chemistry, alter experience. This is not proof, but it’s systematic dependence, which is exactly what emergence predicts…

> The hardest bar would be building a system from physical components, having it report coherent subjective experience So like if i finetune an LLM in a loop to tell you that it is feeling a coherent subjective experience would you accept that? Does that mean that no dog has ever been conscious, because they cannot report a coherent subjective experience? (Because they can’t report anything at all. Being non-verbal.)…

Yeah you’re raising three good points and they all land. On the finetuned LLM: you’re right, that criterion was flawed. A system trained to report experience proves nothing about whether experience is present, which is actually the core of the hard problem. No behavioral output alone can confirm inner experience. That applies to LLMs, and technically to other humans too. On dogs, also a fair correction. We don’t actually require verbal report to attribute consciousness to animals, we use behavioral and physiological evidence. So "coherent verbal report" was too narrow.

Better criterion: a system whose overall architecture and behavior is consistent with experience, not just one that says the right words.

On the standard of proof: that was a rhetorical deflection and you’re right to call it out. You asked a genuine question and got it turned back on you. And you’re pointing at something real: in science, strong correlation is not accepted as proof when stricter evidence is achievable. The reason we settle for correlation here isn’t because it’s sufficient, it’s because subjective experience may make stronger proof structurally inaccessible. But it’s also worth noting that scientific consensus has a poor track record of admitting this honestly. Dominant paradigms tend to defend themselves long past the point where the cracks are visible, physicalism on consciousness is no exception. The confidence with which emergence is presented often reflects institutional momentum as much as evidence.

Re: Nano Banana 2: Google's latest AI image generation model

#542

Post by tmountain »

ACCount37 wrote: post #380
tmountain wrote: post #325

Earlier quoted context omitted.

Art is a reaction to life. AI is thereby incapable of producing anything with any degree of authenticity unless it conveys the experience of being an agent to the world.

What is AI if not "a reaction to life"? With how much data goes into the frontier systems, and how much of it gets captured by them, an AI might have, in many ways, a richer grasp of human experience than the humans themselves do. You were only ever one human. An LLM has skimmed from millions. You have seen a tree, and the AI has seen the forest it stands in.

It’s a subjective conversation but putting AI in the same category as a real artist is like saying someone that’s played a ton of first person shooters has gone to war. It might have a lot of observed information about what is involved in living, but real art comes from a lived experience, just like reading about going to Hawaii doesn’t mean you’ve been to Hawaii. Making something authentic requires synthesizing your life experience with the message you want to convey, and personalizing it in a way that puts an imprint of yourself into the work. Sure, it can render beautiful imagery, but I am speaking to a different issue entirely, and I don’t see any way that it can create in the way I am describing.

Re: Nano Banana 2: Google's latest AI image generation model

#544

Post by imtringued »

avmich wrote: post #269
keiferski wrote: post #237

Some random predictions about what AI image generation tools will do/are doing to art: 1. The narrative/life of the artist becomes a lot more important. The most successful artists are ones that craft a story around their life and art, and don't just create stuff and stop. This will become even more important. 2. Originality matters more than ever. By design, these tools can only copy and mix things that already exis…

I mostly disagree. > 1... The narrative/life of the artist becomes a lot more important. When I watch a movie, I don't care about the artist's life. I care about character life, that's very different. > 2... Originality matters more than ever. By design, these tools can only copy and mix things that already exist. It's like you assigning to humans divine capabilities :) . Hyperbolizing a little, humans also only copy…

>"You're currently here" is the always existing point which isn't achieved at the moment and which gives to critics something to point to and mount the objection to the process as a whole, because, see, this particular thing isn't achieved yet.

This is a contradiction that is so blatant I don't even know what language you're speaking. The definition of that phrase is the exact opposite of what you're saying.

"You're currently here" is the always existing point which is achieved at the moment.

>gives to critics something to point to and mount the objection to the process as a whole, because, see, this particular thing isn't achieved yet.

No it doesn't, because unless progress is reversed or undone, you can always point to your current success and say that the critics have been wrong so far. In fact, that's exactly the argument you're making here, which is why it's so weird that you're twisting it into its opposite.

If you want people to understand you, then you actually have to articulate what you're thinking instead of wrapping it in layers of euphemisms and hoping that the recipient nods along because they happen to agree for a completely irrelevant reason (e.g. "I like AI" or "I like space") to the argument presented.

Re: Nano Banana 2: Google's latest AI image generation model

#545

Post by dockerd »

ghoshbinayak wrote: post #515
kristjansson wrote: post #77

Earlier quoted context omitted.

Part of the job of interior design is delivering the promised images in … yknow, physical reality? How are you going from nano banana images to actual plans, materials, finishes, products, paint codes, … ?

Not everyone can afford the best interior designers, my sister had hired one to design the interiors of her new cafe.. the designer did not provide any renders, just a basic CAD drawing. Then I used nano banana pro to create some renders from the photos of the empty space. These turned out great, and she decided to ditch the designer and use the designs created by nano banana. The cafe has opened now, and it looks gr…

@fasteddie31003 and @ghoshbinayak, Can you share the designs that you created?

Re: Nano Banana 2: Google's latest AI image generation model

#546

Post by keiferski »

chii wrote: post #492
oliyoung wrote: post #410

Earlier quoted context omitted.

> When I watch a movie, I don't care about the artist's life. And here we come back to the aged old "can you seperate an artist from their art" because I'd argue when you watch a movie you are watching a product of their life

The artists life might've been highly affectual and shows in the art, but they doesn't mean the viewer cares about it - at best only so far as it makes the art more enjoyable.

The continual interest in museums, biographies etc. on figures like Van Gogh seem to indicate otherwise. People are very interested in the lives of artists, and without the struggle narrative behind Van Gogh, it’s unclear that he would be famous at all.

Re: Nano Banana 2: Google's latest AI image generation model

#547

Post by mikragor »

buzzerbetrayed wrote: post #496
hirako2000 wrote: post #264

Earlier quoted context omitted.

What about yoldo potato, turn any photo into a vintage looking gem you can print out.

I googled “yoldo potato” with quotes and there was one result. This thread.

Maybe an incorrect translation or an automatic correction? I'm guessing "potato" should be "photo", but who knows. Now I'm curious about that "yoldo potato" thing haha

Re: Nano Banana 2: Google's latest AI image generation model

#548

Post by imtringued »

CamperBob2 wrote: post #437
gambiting wrote: post #411

Earlier quoted context omitted.

>>These entities, whoever they are, they act on our world, they are real, and more and more over time they will get independent from humans, eventually becoming different species that can self-replicate. See, I don't believe that for even one second. They are just very clever calculators, that's all. But they are also dumb like a brick most of the time. It's a pretend intelligence at best.

It's a pretend intelligence at best. The best time to start paying attention was ten years ago, when the first Go grandmaster was defeated by a "pretend intelligence." I sure wish I had. The next best time to start paying attention is now.

I wish I knew what to pay attention to. I've always had trouble with that. I spent 2024 and 2025 learning how neural networks and transformers work. The conclusions of that learning are pretty sobering. Everything uses transformers and despite all the novel architectures that have come out in those years, transformers are still the best and I'm not sure how to come to terms with that.

Does it mean that researchers wasted their time on useless dead end architectures, or are they ahead of the curve and commercial companies are slow to adopt them?

Even the coding agents are more primitive than expected.

Re: Nano Banana 2: Google's latest AI image generation model

#549

Post by anigbrowl »

kevincox wrote: post #52
nickandbro wrote: post #17

These image gen models are getting so advanced and life like that increasingly the general public are being duped into believing AI images are actually real (ex Facebook food images or fake OF models). Don't get me wrong I will enjoy the benefits of using this model for expressing myself better than ever before, but can't help feeling there's something also very insidious about these models too.

I actually think this was a good thing. Manipulating images incredibly convincingly was already possible but the cost was high (many hours of highly skilled work). So many people assumed that most images they were seeing were "authentic" without much consideration. By making these fake images ubiquitous we are forcing people to quickly learn that they can't believe what they see on the internet and tracking down sour…

And if they don't?

Your post seems a little naive to me, a lot of people are just not interested in putting in the work or confronting their own confirmation bias, and there's an oversupply of bad actors who will deliberately generate fake imagery for either deception or exhaustion. Many people are just not on quest for truth and are more interested in the activation potential of images or allegations than in the factual reliability.

Re: Nano Banana 2: Google's latest AI image generation model

#550

Post by ralfd »

rootusrootus wrote: post #203
manuelabeledo wrote: post #110

Earlier quoted context omitted.

> By making these fake images ubiquitous we are forcing people to quickly learn that they can't believe what they see on the internet and tracking down sources and deciding who you trust is critically important. Has this thought process ever worked in real life? I know plenty of seniors who still believe everything that comes out of Facebook, be AI or not, and before that it was the TV, radio, newspapers, etc. Most p…

> I know plenty of seniors And not just seniors. I see people of all ages who are perfectly happy to accept artificially generated images and video so long as it plays to their existing biases. My impression is that the majority of humanity is not very skeptical by default, and unwilling to learn.

Yes. People willingly accept made up text (stories) if it fits their world view, and for words we always knew that they could be untrue. Why should it be different for images/audio/video?
Post reply on HN