I don’t think you should respond to questions about your credibility by telling your interlocutor not to ask the question, with the implied point being that the underlying nobility of your quest is so unimpeachable that your credibility is without question.
If your AI characters sound fake, the fix is almost always in how you write and DIRECT the dialogue in your prompt.
I recently started on the dialogue-heavy scenes for my AI feature film about Beethoven, and too many of the lines kept coming out wrong.
They’d sound too theatrical, too monotone, or too much like a practice read-through between actors.
(all this despite the fact that Seedance 2.5 is fantastic when it comes to realism.)
So through a whole lot of trial and error, I've found that these 5 things made the biggest impact on the quality of the dialogue delivery in my generations:
1. Paying close attention to the written punctuation used.
2. Breaking long lines into shorter ones, with intentional pauses.
3. Using bodily actions as punctuation (this one gave the biggest improvement BY FAR! In the attached clip, note the sniffles in between Beethoven's lines. That alone made the delivery feel much more natural.)
4. Using "anchor gestures" (he lifts his finger as if to say "one sec")
5. Just split your generations in two (stupid simple, but it really makes a difference.)
So if your characters keep sounding like they're reading from a teleprompter, give this a try!
@i_ikhatri I find this claim always very bizarre. The reality is that the first tier AI people basically all buy the safety case, and the people who were never close to anything actually important grandstand about how they have way too much experience with AI to accept it.
announcement: i'm launching a $100,000 grant for amateur astronomy research: https://t.co/QQIpI7ink8
over the past few weeks, @joshuawolk@LukeFarritor and i have been digging through public telescope data, hunting for new planets, asteroids, and more
our first release: we found a new dwarf planet in our solar system!! it's insane how much you can find with some claude + codex credits
it's about 1/4 the width of Pluto and orbits about twice as far (way past Neptune), and looks like it's about the 50th largest object in our solar system. @joshuawolk is in the process of confirming it
this dwarf planet has sat in public images for over a decade without anyone noticing. there's tons more out there: all sitting in public NASA/ESA/etc. databases for you to analyze
there's tons of stuff still out there, sitting in data that astronomers don't have enough time to go through all of
you can do this too! apply for a grant. more info below
All hail the weird web
I just enabled HTML-in-Canvas in Claude’s desktop apps. It’s a footnote change next to everything else, but with it, Claude can make bizarre & beautiful sites. Reminds me of the Flash days!
Please look at the stupid demos I made: https://t.co/kNLJoRwmje
it should not be illegal in the US to have a political position regarding national security opposed to that of some government officials and actively lobby for it
and export controls should not violate the principle of no ex post facto enforcement
It took about a full week and a ridiculous number of Fable credits, but Claude figured out how to train a 6M parameter model that fits into 3.2 MB optimized for Game Boy. The result: on-cartridge generative conversation with a finch oddly knowledgeable about the Revolutionary War
absolute fanfic. AI has real, immediate, tangible dangers that we need to focus on, for instance the impact on the job market and making sure random script kiddie can't use it for cyber offense.
But this "oh once it's superintelligent it will take over and ensave us, maybe kill us all!" is pure sci-fi fanfic and distracts from the real (and difficult) issues, which linear probing won't solve.
It's worth repeating the LLMs don't just "decide" to do something and then just "do it" in a void. They generate tokens based on our prompts, only as long as we call `model.forward()`, and on giant clusters with terabytes of weights, that's impossible to "quickly self-replicate".
We also need to regularly remind that the OAI/HF incident was specifically a _guardrail-free_ cyber eval where the model is _instructed_ to hack, to find out how good it is at it.
Some recent personal event made me realize there needs to be more push-back on this doomsday nonsense, and we need to re-focus the discussion on real dangers/issues. I won't argue much in the replies though, not interested in peer-reviewing Terminator fanfic.
I’m very glad for open source. It wouldn’t have been possible for @nickcammarata et al to do their excellent mechinterp research without a modern Qwen model. I hope the offense defense balance is such that we can have open source models for a long time
I like Theo generally, but this sort of bad faith gotcha attitude towards critiques of technology that are so prominent in the Valley really needs to stop. You know this is not what he means, and the arrogance and hubris that bleeds through this is part of the reason we struggle to reach the rest of the world! If we are to do good for humanity by technology, we should be far humbler.
This is the craziest improvement in AI ever, only a year ago AI with threejs produced literally laughable result, now it’s basically a image and video model
I’m just blown away when compared side by side. By this time next year I can’t even imagine how awesome it would be🤯
the chinese have rightfully earned enormous global goodwill and soft power with their open source model efforts. even most Americans in this space feel more sympathetic to them than with the model companies