"... If you don’t own the query interface, you’re just assembling training data for those who do."
Good point, but Google is already there.
"A highly personalized AI-powered medical assistant will complement (and eventually replace) our General Practitioner."
No way. Chatbot-type AI has no notion of what it's doing or the implications of its advice. We're several major breakthroughs from that. Although someone may try this for some low-rent level of health care.
"We can no longer rely on our senses to determine truth."
We can no longer rely on imagery, video, or audio. Everything now needs a full chain of custody to be trusted. Which we will not often get.
"Hyper-personalized experiences disrupt traditional e-commerce and hospitality brands – and enable an immersive era that is tailored to you."
Another filter bubble scheme. We already have personalized ads. We could have systems where you walk into a chain restaurant and they already know what you want. Hotels might do something for frequent travelers. But they could have been doing that ten years ago.
"Companies will perform rapid resets to eliminate organizational debt and compete with startups."
Historically, this rarely happens, even when it needs to. When it does, it usually involves firing the CEO.
Historically, this rarely happens, even when it needs to. When it does, it usually involves firing the CEO.
Scott Belsky is an executive, entrepreneur, author, and investor (and all-around product obsessive). He currently serves as Adobe's Chief Product Officer and Executive Vice President, Creative Cloud. Scott's passion is to make the creative world more productive, connected, and adaptive to new technologies. Scott co-founded Behance in 2006, and served as CEO until Adobe acquired Behance in 2012. Millions of people use Behance to display their portfolios, as well as track and find top talent across the creative industries.
I think you're calling this one correctly, this is what passes for shopping your resume around town at that level.
Just click the uBlock icon, select the "element picker" and click on the element you want to remove. You will get a preview and when you are satisfied it will be saved to your preferences and you will never encounter that popup again.
You could also click the "</>" icon instead to disable Javascript for the site. Substack has subdomains for every site though so you want to click the "dashboard" instead, go to "My rules" and enter "no-scripting: substack.com true" to disable it for all their subdomains. It used to show an irritating please turn on javascript-message, which can be removed in "My filters" as "substack.com###nojs-banner".
It’s just me, but I do not call anything without a clear time line a „forecast“. Is it a forecast if you cannot decide clearly if it has happened or not? What then have you forecast?
Here’s something I don’t get. If the generative AI gets so good it can diagnose disease or give us perfectly curated experiences, why wouldn’t it just “generate content” at an expert level? Why would we need to compensate “content creators” at all. It would just take our preferences and give us an endless stream of Avengers themed content (for example), generating new stories, blog posts, and fan sites as fast as we can consume. For that matter, it could give us expert level prediction articles as fast as we can read them.
I'm waiting, but don't believe it is/will be "that good". Language models are about language, not knowledge - our human gullibility and hopefulness is a vital factor in the current "success" of generative systems.
I was actually trying to point out the irony of the idea that artists and programmers should get used to being displaced by AI, while this author won't?
But that being said, I could see a world where the language model AI generates new Avengers scripts, while the deep fake AI and the CGI AI generate the movie. The reviewer AI tailors the review to the reader, on demand, so the reader is assured that their choice of entertainment is awesome! Then the blogger language model AI goes into all the possible plot twists and easter eggs that may have been in there, generating endless YouTube fan videos.
We may, one day, live in a world where fandom gets 1 minute of new content every 30 seconds. Maybe completely on-demand so that, yes, the real hero is a hypertensive, 42 year old, who was pivotal to helping the super-hero hero out! And as long as it's canon and pretty much like the other stuff that same hypertensive, 42 year old likes, it's great! Did you know it's Canon that Aquaman is on Lipitor and drives a Buick? I should totally ask my doctor about a Buick... or is it the Lipitor?
From the ‘product obsessive’ that brought the world Adobe’s creative cloud - a product set that helped move forward the subscription economy and was roundly panned.
"Highly personalized, AI-powered medical assistant will complement (and eventually replace) our General Practitioner."
Doubt it. Regulations and jobs protectionism will make this illegal to do. Lower regulated fields with less powerful incumbent interests will be disrupted first.
"The era in which we can no longer believe our eyes ushers in “The Age of Attribution” and a default distrust of our human senses."
Doubt it. The human species has a tendency towards arrogance, narcissism and away from introspection, humility. The default will remain belief in one's own initial perceptions and judgement.
20 comments
[ 9.7 ms ] story [ 66.3 ms ] threadGood point, but Google is already there.
"A highly personalized AI-powered medical assistant will complement (and eventually replace) our General Practitioner."
No way. Chatbot-type AI has no notion of what it's doing or the implications of its advice. We're several major breakthroughs from that. Although someone may try this for some low-rent level of health care.
"We can no longer rely on our senses to determine truth."
We can no longer rely on imagery, video, or audio. Everything now needs a full chain of custody to be trusted. Which we will not often get.
"Hyper-personalized experiences disrupt traditional e-commerce and hospitality brands – and enable an immersive era that is tailored to you."
Another filter bubble scheme. We already have personalized ads. We could have systems where you walk into a chain restaurant and they already know what you want. Hotels might do something for frequent travelers. But they could have been doing that ten years ago.
"Companies will perform rapid resets to eliminate organizational debt and compete with startups."
Historically, this rarely happens, even when it needs to. When it does, it usually involves firing the CEO.
Scott Belsky is an executive, entrepreneur, author, and investor (and all-around product obsessive). He currently serves as Adobe's Chief Product Officer and Executive Vice President, Creative Cloud. Scott's passion is to make the creative world more productive, connected, and adaptive to new technologies. Scott co-founded Behance in 2006, and served as CEO until Adobe acquired Behance in 2012. Millions of people use Behance to display their portfolios, as well as track and find top talent across the creative industries.
I think you're calling this one correctly, this is what passes for shopping your resume around town at that level.
Actually, that just removes the floating header, which probably isn't what you want.
You could also click the "</>" icon instead to disable Javascript for the site. Substack has subdomains for every site though so you want to click the "dashboard" instead, go to "My rules" and enter "no-scripting: substack.com true" to disable it for all their subdomains. It used to show an irritating please turn on javascript-message, which can be removed in "My filters" as "substack.com###nojs-banner".
Create a stylesheet for substack.com.
Add:
Edit: s/prediction/forecast/a
I'm waiting, but don't believe it is/will be "that good". Language models are about language, not knowledge - our human gullibility and hopefulness is a vital factor in the current "success" of generative systems.
But that being said, I could see a world where the language model AI generates new Avengers scripts, while the deep fake AI and the CGI AI generate the movie. The reviewer AI tailors the review to the reader, on demand, so the reader is assured that their choice of entertainment is awesome! Then the blogger language model AI goes into all the possible plot twists and easter eggs that may have been in there, generating endless YouTube fan videos.
We may, one day, live in a world where fandom gets 1 minute of new content every 30 seconds. Maybe completely on-demand so that, yes, the real hero is a hypertensive, 42 year old, who was pivotal to helping the super-hero hero out! And as long as it's canon and pretty much like the other stuff that same hypertensive, 42 year old likes, it's great! Did you know it's Canon that Aquaman is on Lipitor and drives a Buick? I should totally ask my doctor about a Buick... or is it the Lipitor?