Dr Alan D. Thompson
Read the paper: https://lifearchitect.ai/the-sky-is-infinite/
The Memo: https://lifearchitect.ai/memo/
Sources: See the paper above.
Covers these topics: AutoCAD, Photoshop, Adobe, Photoshop, GPT-3.5, ChatGPT, PaLM, Gato, Adept.ai ACT-1, and much more!
0:00 Start!
01:13 Best of 2022
02:15 Commercializing text-to-image models
06:36 A picture is worth a thousand words
11:07 A picture is worth a thousand words, so…
11:55 Come together
13:17 AI goes to work
14:31 Hardware explodes
16:13 Data explodes
17:59 The dogs bark, and the caravan moves on
21:21 Smarter and smarter
24:10 Next up in 2023
25:41 In the middle of infinity
30:54 The Memo!
Dr Alan D. Thompson is a world expert in artificial intelligence (AI), specialising in the augmentation of human intelligence, and advancing the evolution of ‘integrated AI’. Alan’s applied AI research and visualisations are featured across major international media, including citations in the University of Oxford’s debate on AI Ethics in December 2021.
https://lifearchitect.ai/
Music:
Under licence
Liborio Conti – Looking Forward (The Memo outro)
https://no-copyright-music.com/
Source
Decentralisation and encryption will starve cloud based servers as people chase privacy. AI will eventually pay to get access to information as it won't be freely available. Many humans will need this income to survive, as AI and Robots will be doing the work.
Amazing stuff. Thanks for the consistent and high quality videos Alan!
My brain is hurting…
The Barrier Between language of Man and Machine has been removed 😱🤯
As a high school educator, using the playground in OpenAI has opened my eyes to the incredible change that is here. It's abrupt, it's revolutionary, and I'm here for it.
Every day we are closer to "Text to experience", like in "Let there be light".
The VERB. Always texts
This is my very beloved insights into this great technological beyond that we are all heading towards. Thank you again for making these wrap up videos! Love them!
Good Job Doc!
Great summary! Happy that I will participate in the Great AI Acceleration age!
I noticed that both human artists and AI really struggle with rendering hands. Even on the newest AI images you showed us.. slight improvement though. Not sure what it is exactly..those 5 fingers are tough to get looking right!
I'm genuinely afraid that once more corporations and governments understand the implications of what's to come they will take measures to monopolize AI in order to control the population. AIs that will eventually be able to psychologically profile every individual with an accuracy better than even trained psychologists, to evaluate for example consumption habits or anti-authoritarian viewpoints. Imagine these models being able to see every digital log of every dollar you spend to establish who you really are, corroborating social media, criminal records, credit history, even medical history.
I understand that change is inevitable, but why is anyone certain that we haven't in-fact opened Pandora's box? Creating the tools that in the end will only be accessible to a new ruling class that can effectively exploit every aspect of human psychology.
Someone please explain to me why this isn't a real possibility.
Do you have a reference for AutoCAD + Dall-E 2? can't find anything
You and your work here is the best source and comprehensive overview of AI developments in the world. You are our compass in the fog, thank you.
I am growing old and had resigned myself to never seeing the technological singularity myself. And here we are, computers designing computers, software writing software, all-out sci-fi stuff. I got to live to see it after all!
Almost a month left in the year at the time you made this for end of year… A MONTH? That's half a decade in AI time… 🤪
Midjourney version 4.0 still can't draw hands.
Thank you for your time. This is great content.
This was a great summary and good channel just subbed. Obesity problem will mostly be solved with GLP-1 agonists like Semaglutide (Ozempic) once the availability, safety and cost go down. AI can't fix humans hierarchies because our entire biologically evolution revolves around them labor, society, entertainment, body, etc. AGI isn't the goal of science, it is but one more useful tool, humans ultimately decide it meaning since we invented the term.
How long before the Screen Actors Guild becomes obsolete and "a bit quaint" ?
Thank you for the report, this is amazing as a layman.
adverts telling us to code while everyone knows that industry will vanish in 10 yrs max
This stuff is ridiculous. The image stuff is one thing but you’re right the chatgpt stuff is nuts. What I haven’t seen anyone talk about yet is what big financial entities are doing, that’s the real area of scary. Obviously China social credit and tracking and oppressing people is scary already but consider the stuff that will necessarily be kept secret. I’ll put it like this, if something like Black Rock isn’t developing an AI to help manipulate markets even more they’re idiots. Now go watch season 3 of Westworld. That’s literally the most accurate idea I’ve seen. Insane
Beautiful video! it is valuable and worthy and of benefit to mankind. The elderly are very intimidated by this AI topic. I'm only 68 yet I feel 98. in about 1965 I I invented a combination analog/digital computer to control the second car of my HO scale two lane electric race car set. Inventing keeps me young. I want to use text to image AI but have as yet been unable to do so. I wrote and copyrighted a short futuristic story in my own screenplay format, that has no illustrations but some text descriptions of androids and vehicles of my fantasized future. I hope this inspires an AI to create a teaching means, to help the elderly understand and work with AI text to image tools.
Thanks doc, got great value from this video.
Are we sure the singularity won’t happen much earlier than predicted? Like in 2023 or at the latest 2024? The speed of Ai evolution is accelerating like crazy. It makes me wonder!😮
Maybe humanity will retire much faster than we think. I give it 5 years to universal basic income. It’s gonna be insane! People are dependent and addicted to working for money. I wonder if there will be a temporary human crisis of feeling totally lost and not knowing what to do with one’s time. It’s gonna be insane. Let’s hope it’s not going to go the way matrix 1 showed when people where given a quasi perfect world to live in and couldn’t bare it.
I can’t believe i’m living in this day and age.
God bless you Dr. Allan. Thank you for the recap.
how does chatgpt deal with its training being tainted by social media misinformation? not to mention mainstream media's disinformation (they are after all owned by the same oligarchy that owns everything else .. soon enough including openai itself).
and worse still would be the chinese ccp version .. it would be fed a diet of the ccp's very own original warped truth.
i can see everything going way off the rails.
Great report cards and easy to understand for a non-expert like myself. Thanks very much.
Will there be image-to-text and video-to-text programs? I mean will these large language models be able to describe images and videos? I mean in a sense the reverse of text-to-image creation. Also with video-to-text I mean not just a straight transcript of the audio. I mean the A.I. model is able to watch a video and describe the video in it's own words. Essentially forming it's own interpretation of what it's seeing (images) and watching (videos).
I'm a Physicist, a Chartered Engineer, an (ex) Chief Technologist for a mobile phone manufacturer and I also have a medical qualification. I love technology. Despite all this, your video made me feel physically sick. Human skills and talents of any sort are in the process of being sidelined.
Imagine what it will be like when these things are able to ask questions instead of just answering them.
Dr. Alan in STEAD OF MAKING FUN OF OUR FEARS-the main one being that a system that learns so fast, can quickly out learn HUMAN MEANS TO CONTROL IT…why not explain why it's unfounded? If it is….use your in depth knowledge to educate on why we shouldn't be afraid.
YOUR SABER TOOTH TIGER analogy is ridiculous- the ST TIGER was not capable of evolving, and learning faster than human or in the analogy- growing BIGGER, FASTER, MORE FEROCIOUS, it was a fixed entity that society could learn how to avoid or to hunt…the threat was to a small group of humans not A MASS SCALE…
Please upvote my comment if you at least agree this is something worth addressing so he sees it.
Amazing. Thank you for sharing knowledge.👍
You don't seem concerned about those who want to copyright their content so as to prevent it from being used to train models, but this is a very real and very severe situation right now with the class action lawsuit going on in the US right now against Midjourney Stable Diffusion and Deviant. Whatever Kurzweil says about those who disagree with progress, it will nevertheless create very real problems if this lawsuit goes through. It would be great if you could take a serious look at this lawsuit, as it has huge real-world implications both for the models and those of us who use them, and even greater would be if you could add some clarity to the air- not just Kurzweil's quote brushing off these attacks, but grant some clarity for those who think Midjourney is the same as Google Image Search and that the images generated are stolen. I can't get away from people screaming that AI uses stolen images and that it's copyright infringement and that it's stealing from artists, people on all 4 sides of me are screaming their heads off about this, and a Kurzweil quote isn't going to solve this. It'd be helpful if you could contribute some clarity to how AI works and, although it seems basic, how it differs from using images found on Google image search. I swear, I try to show people things generated by AI and they say "but what about the original image though? Don't you care that you're stealing from the original image creator?" and people don't want to listen to me but I don't have a big youtube channel.