Emad Mostaque is a visionary, genius but also a generous man. Open sourcing the model is a such an unprecendented move. It is indeed true that the benefits far outweigh the benefits. Having large tech companies be arbiters of the truth is outright stupidity indeed. Thanks again Emad and your team. I would have liked to work for you if I had the option to quit my job. You're doing God's work bro.
I too am genuinely impressed by him and how he is thinking about the big picture -- what this means for humanity. I think the window to make sure that his vision is the reality is quite small though.
My biggest fear is of other nature: that today's startups in the field will become future mongrels suprassing Google, Microsoft, Meta, etc., and the initial selling points of accessibility and democratization will just become the thing of the past. Like a nice inspirational quote you hang on the wall. The generative AI field is so disruptive that, if not followed by deep, pro-equality socio-economic changes in the world, will just make majority of people's life miserable. Very nice and thorugh interview all and all, Nina.
Quite an exiting development! A quick question though. I havent' heard much about the issue of energy consumption and how massive adoption of AI technology could eventually impact our global energy consumption. Is there any discussion available that delves into this question? Is this a warranted interrogation? With such volotality and uncertainty in the energy market, I guess that some AI companies, researchers or investors have touched upon the issue. Would be curious to know their thoughts. Thanks a lot for this first content, looking forward for the following ones!
I would be very interested in an open debate between somebody advocating complete open-source access to this technology and somebody advocating a much more closed, safe-guarded approach. It is not clear at all to me which is better in this instance, and it would probably be better to figure it out before any bad consequences occur. Emad vs Altman or something would be interesting, lots of options
I mean, he recognizes the huge PROBLEM of "democratization" when he says at 00:31:57 "I mean this policy for example so in the European Union right now there is legislative pressure being pushed through such that General AI models um the creator of the model is liable for the end use just like as a financial institution you're liable to the use of your platform um that flips it such that basically it banned open source AI in Europe", but his adoration of the myth of "democratization" leaves him like a helpless infant, incapable of defending himself against "democratiization", which is nothing but a modernized "might makes right" form of old-world "right to rule", which isn't actually a right.
25:08 So, do I understand this correctly? Artists contacted Stability AI after finding their images in the Open AI/Clip dataset. Which would, if Stability AI didn't already know about it, be a concerning discovery. Yet Stability AI then proceeds to merge the Open Ai model/dataset with LAION, to create their latest model. I'm sorry, that just doesn't seem like the right way to do things.
Hey. No, he’s saying that the first version of SD had clip in it - a model from Open AI. When artists contacted Stability about why they were in the dataset, they figured out it was because of Clip. So they removed Clip from V2 - and asked LAION to build a new dataset instead.
Revelation 13:15 “And he had power to give life unto the image of the beast, that the image of the beast should both speak, and cause that as many as would not worship the image of the beast should be killed.” Transhumanists do Not write The Future
This model mirrors the sports model where everybody has access but you have to move up the engagement hierarchy to get better access. In other words, your skills and applications need to resonate better with the larger conversation for you to access better tools.
Im surprised that Emad seems to be completely unaware of the Surveillance Capitalism phenomenon that has captured the motivations of the worlds largest corporations and more. In other words, its NOT ONLY about selling ads.
When you encounter controlling "interest" from the FBI, CIA et al, will this skew your agenda, as has happened in all big media orgs? Do you have an idea of how you may address that inevitability?
It's funny that he says "if you think unelected individuals are the right arbiter you're stupid" because he thinks elections prevent extremism. "Majority rule" ideology is the positive feedback loop that AMPLIFIES extremism.
@Ravi you're mixing up Stable Diffusion and OpenAI these are different companies working on different models, the former with CEO here is focused currently on image generation (text to image), whereas ChatGPT is a text to text transfomer model.
As long as you don't work with GPT all is well.. The problem is that the GPT architecture is basically a mimic of the brain. The Generative stands for the brain stem were basic text operations are done, the Pretrained network is similar to the thalamus and the Transformer layer is similar to the Cerebrum. So if you are doing GPT you are actually producing a person. Please don't.
@@CyberwizardProductions he used the word Jihad (to label those terrorists) which is one of the core fundamental concepts in the context of islamic just war theory
Nina, you have really set a high bar with Emad for your first interview. Looking forward to your future podcasts.
That's so kind of you! More interviews dropping soon!
Emad Mostaque is a visionary, genius but also a generous man. Open sourcing the model is a such an unprecendented move. It is indeed true that the benefits far outweigh the benefits. Having large tech companies be arbiters of the truth is outright stupidity indeed. Thanks again Emad and your team. I would have liked to work for you if I had the option to quit my job. You're doing God's work bro.
Very uplifting! A truly wonderful step in the direction of the Open Inclusive Future. A generous gesture of profound good faith.
Amazing interview!
A truly engaging conversation. Thanks
great interview
Thank you!
Great content! Keep it up n thanks!
everyone share this out on every social media outlet you can. And get your followers to share it out, too
Thanks for share
Emad you rock! Thanks Nina
Emad Mostaque is the man of the people. Love this guy.
I too am genuinely impressed by him and how he is thinking about the big picture -- what this means for humanity. I think the window to make sure that his vision is the reality is quite small though.
My biggest fear is of other nature: that today's startups in the field will become future mongrels suprassing Google, Microsoft, Meta, etc., and the initial selling points of accessibility and democratization will just become the thing of the past. Like a nice inspirational quote you hang on the wall. The generative AI field is so disruptive that, if not followed by deep, pro-equality socio-economic changes in the world, will just make majority of people's life miserable.
Very nice and thorugh interview all and all, Nina.
I am honoured to be one of the first to discover this channel!
Great interview. Well done!
How on earth did you get Emad on your tiniest of channels? Keep the good interviews coming! Subscribed!
great discussion and insights
great conversation 🤌🤌 .... wouldn't be boring if you two would continue such conversations from time to time ...
great interview
Quite an exiting development!
A quick question though. I havent' heard much about the issue of energy consumption and how massive adoption of AI technology could eventually impact our global energy consumption.
Is there any discussion available that delves into this question? Is this a warranted interrogation?
With such volotality and uncertainty in the energy market, I guess that some AI companies, researchers or investors have touched upon the issue. Would be curious to know their thoughts.
Thanks a lot for this first content, looking forward for the following ones!
boring
great insights, and when was this session recorded.
I would be very interested in an open debate between somebody advocating complete open-source access to this technology and somebody advocating a much more closed, safe-guarded approach. It is not clear at all to me which is better in this instance, and it would probably be better to figure it out before any bad consequences occur. Emad vs Altman or something would be interesting, lots of options
Never heard of you. This is mind blowing. Very insightful questions! Got my sub.
Awesome, thank you!
Excellent!😁
Great discussion with Emad.
Really excited for this new phase of the human journey.
How can we read your work?
Sign up to my Substack! ninaschick.substack.com
I mean, he recognizes the huge PROBLEM of "democratization" when he says at 00:31:57 "I mean this policy for example so in the European Union right now there is legislative pressure being pushed through such that General AI models um the creator of the model is liable for the end use just like as a financial institution you're liable to the use of your platform um that flips it such that basically it banned open source AI in Europe", but his adoration of the myth of "democratization" leaves him like a helpless infant, incapable of defending himself against "democratiization", which is nothing but a modernized "might makes right" form of old-world "right to rule", which isn't actually a right.
25:08 So, do I understand this correctly?
Artists contacted Stability AI after finding their images in the Open AI/Clip dataset. Which would, if Stability AI didn't already know about it, be a concerning discovery. Yet Stability AI then proceeds to merge the Open Ai model/dataset with LAION, to create their latest model.
I'm sorry, that just doesn't seem like the right way to do things.
Hey. No, he’s saying that the first version of SD had clip in it - a model from Open AI. When artists contacted Stability about why they were in the dataset, they figured out it was because of Clip. So they removed Clip from V2 - and asked LAION to build a new dataset instead.
@@ninadschick Thanks for the quick answer.
Ok, I see. When he said "both of them are LAION" I thought he was referring back to Clip.
Revelation 13:15
“And he had power to give life unto the image of the beast, that the image of the beast should both speak, and cause that as many as would not worship the image of the beast should be killed.”
Transhumanists do Not write The Future
This model mirrors the sports model where everybody has access but you have to move up the engagement hierarchy to get better access. In other words, your skills and applications need to resonate better with the larger conversation for you to access better tools.
Im surprised that Emad seems to be completely unaware of the Surveillance Capitalism phenomenon that has captured the motivations of the worlds largest corporations and more. In other words, its NOT ONLY about selling ads.
Shoutout to Replit
When you encounter controlling "interest" from the FBI, CIA et al, will this skew your agenda, as has happened in all big media orgs? Do you have an idea of how you may address that inevitability?
I very much like Emad's emphasis on open intelligence....if it's not open, its not intelligent, just another closed ego structure.
It's funny that he says "if you think unelected individuals are the right arbiter you're stupid" because he thinks elections prevent extremism. "Majority rule" ideology is the positive feedback loop that AMPLIFIES extremism.
now that MSFT has invested $1 B in OPEN AI, it is no longer an non-profit...and MSFT is also in cahoots with US Govt.
Open AI was never truly open.
@Ravi you're mixing up Stable Diffusion and OpenAI these are different companies working on different models, the former with CEO here is focused currently on image generation (text to image), whereas ChatGPT is a text to text transfomer model.
As long as you don't work with GPT all is well.. The problem is that the GPT architecture is basically a mimic of the brain. The Generative stands for the brain stem were basic text operations are done, the Pretrained network is similar to the thalamus and the Transformer layer is similar to the Cerebrum. So if you are doing GPT you are actually producing a person. Please don't.
13:57 as someone who fights against islamophobia, i found it discouraging that he is associating terrorists with anything islamic.
re-watch that, I think you missed what he was actually saying. he wasn't associating it with anything islamic.
@@CyberwizardProductions he used the word Jihad (to label those terrorists) which is one of the core fundamental concepts in the context of islamic just war theory
Emad is the fool within the tarot