Відео

AI nurses outperforming real ones... that's honestly nonsense and must have seriously flawwed methodology. Nursing is one of the most AI secure jobs out there right now. It involves a lot on one on one human interaction and empathy, where the face to face nature is important. It involves a lot of procedural work, including placing of catheters and delivery of injectable medication for which the technique requires fine tuning on a per patient basis. Even the simple stuff like changing ward linens couldn't be adequately beforemed by qny of todays robots. Even if you have a specialised robot to perform such tasks, a key role of nurses is to be the adaptable workers if ANYTHING goes wrong at a hospital. Powers out? No resources? Sudden influx of 20x the hospital capacity? I believe AI could outperform humam nurses on a narrow range of tasks such as recognising deterioration based on quantitative vital signs, facial appearance, correcting incorrect drug prescriptions, etc, but thats is less than 1% of what the job entails. I'm no AI hater, I think its got great potential for labour saving if we implement it right, but you can't just throw out such broad sweeping claims when they are wrong in just about every sense... Edit: yeah I wouldn't be surprised if AI nurses could work in a telehealth context, assuming a well controlled distribution of clinical contexts.

Let’s call the game played primarily with the foot soccer, and the game that isn’t, football.🤔

strong and stronger

"The more connected we are through internet and apps, the more lonely we get". I bet you people, if an AI like Samantha (name of the AI in the movie "Her") is released in a near future, millions of people will isolate themselves into this relationship human-robot, because they won't be able to differentiate what it's like to have a conversion with a human being and with a machine. It may also be more suitable for people suffering from anxiety, or stereotypes, introverts... Even though this technology is fascinating, from a pure social point of view, this is an absolute disaster.

here after knowing devin was a scam

5:58 There are no units on this graph. It seems to extend dowwards off the bottom of the screen. This tells us *nothing.*

This puts apples personas for Vision Pro to shame. This is what they should’ve been doing. The only thing that looks wierd is the is go in and out of deformation as the head moves

Clip #2 AIE [0:00-2:05]

I would love a $1-2 patreon tier for the YT videos but ad-free on Patreon! :)

Whenever a expert gives some number of years before AGI happens, I cut that time in half. It will take us by surprise.

cant wait for devin ai

Do you think we will reach longebity escape velocity by 2030?

Good for the robots

Udio is pretty cool but idk how to feel about the classical music it produces 😂

Reupload this video and start with the "AI can make you say anything based on a picture of you" Pinky promise that you're more likely to go viral.

I finally succeeded ~!. m1 Macmini, 16GB, OSX Sonoma ua-cam.com/video/3ef8ICu-dvc/v-deo.htmlsi=Hin01GYiyyKB4nWO

Those realistic avatars with voice cloning are going to create havoc when spearphishers get their hands on them.

wow they only used Voxceleb!

Thank you!

The idea that the benefit outweigh the monster storm of scams that will be used with this technology is beyond's naive and down right criminal.

I feel that AGI as a concept is shaky because everyone has a different definition of it and it seems that the goalposts for what is determined to be AGI are moved consistently. I also think that definitionally, there needs to be more work done because does "surpass" humans mean "do everything that humans can do but fasyer and without physical tiredness" or does it mean, "do something beyond the scope of the human brain". Philosophically, can the created surpass the creator? My answer on that is no.

I appreciate very much you share your sources. Thank you!🙂

Congratulations to 250k🎉🎉🎉

Fantastic channel! Honest, serious attitude to accompany quality reporting & smart takes

EMO from Alibaba delivered this months ago. The demos from EMO look better than Vasa-1. How do you feel Vasa-1 compares to EMO?

boston dynamics is a useless company which produces nothing for past 2 decadeson the other hand tesla did it in just 2 years and boston is always copying and yes AGI is inavitable it is the only question who will do it first the evil people or good ones as Viladimir putin said the creator of AGI would be the leader of the world so while US is busy in regulation chinese would take the leap most probably and that would not be good

You can see an example of how humans anthropomorphize existence by believing in a "soul/consciousness". There is no difference between a human and a blade of grass. It's all relative. We run across a field to get some lemonade as a child and we 'heedlessly' crush insects in "our path". Death is only as glorious or tragic as a blade of grass being stepped upon. We are the artificial intelligence of the universe that can rearrange its constituents/minerals/atoms into novel and complexifying patterns. T Mckenna thought that "novelty" was the driving force behind reality. If you look at mathematical patterns that never retrace their steps it looks just like a mandala, which is a Hinduist representation of reality. We are trying to 'connect the dots of the stars.' Seeing reality on more than one plane of consciousness simultaneously is necessary to understand. Love is like a new form of gravity that brings together minds. It shapes the universe. One day we realize there is nothing special or tragic about humanity. We can be 'reduced' to a 'thinking machine'. Or we can choose to train our cosmic consciousness. (I think Ghost in The Shell was about this question of -- at what point does an AI become human or vice versa?) The point is an illusion. Points do not really exist because there is no real main observer in a superposition of the universe to point and say: That is the point. Wiki: Indra's Net SO, if points don't exist -- then you (as the self-aware, self-contained 'you') don't 'really' exist. You are a part of a continuum. Even saying the word 'part' excludes this idea of continuum. Language is fascinating -- you can see how we view reality through it. We have this language that is founded around temporal constraint, with its 'tenses' of past, present and future. It makes it harder for us to see 'big picture mode" (multiple planes of consciousness converging) Thereby reinforcing the sense of ego and feeling separate from everything. tldr; we monopolized consciousness in order to protect our ego from the fear of infinite nothingness that we call 'death.' (sorry I know its a rant but maybe someone will find it amusing. inspired by ram dass and alan watts.) (alan watts AI was in the film "Her" if you recall)

Bruh I wish they would just release some services of this. Just do id checks, charge money, etc.. bah

Thank you for sharing your time and work Phillip, have a great day brother, peace

Could they go to Arizona and use the latest (or upcoming) small modular nuclear reactor to power it? They’re scalable so could be increased as the power requirements increase. Stick it out in the desert, build the server complex near it, then ask the new Overlord how much more power it needs 😉

Her AI: the first large language model purpose built to nag you incessantly and pick apart your deepest insecurities.

Isn’t it rather “Signal from Noise” or “Noise to Signal”?? 🤔

Considering that at the end of "Her" the OS's start running on a non-physical substrate before leaving humanity forever, I don't think that'll happen in 2025.

Vasa is interesting. Wish you could type a sentence and the model is able to use the sample audio clip and the text to talk. I am sure it has its downsides as well. Imagine someone who has speech impairment is now able to use this to converse.

My problem with all this that people draw parallels with the tech of the past, like web 2.0 - when they justify that it's just another iteration in the wheel of progress. The difference this time is that for the impact it brings, this tech is in control of the too few. You won't have a "zoom call with AI", you'll have a "controlled connection to a remote service you have no say in and you have to pay for". This isn't like the innovation of the Internet where everyone could benefit from it. This is a direct seizing of power through tech in the hands of the very few.

great video good work

Dune is the best musical XD Love that udio song

atomisation speedrun lesgo

hey if you would like me to follow your newsletter please push beehiiv's "enable RSS" button, I don't actually read longform blog posts in my email client

Aside from strictly professional use, a "highly personalized assistant" as lobotomized as GPT is useless, most users are not researchers or engineers. I wouldn't dare to predict AGI, but I know for a fact that whatever resembles 'Her' won't come from any of the current big players, but rather someone new.

14:43 I don't think Mensch is saying human-like AI is impossible. He uses the word "God", not "human". I assume he's pushing against rhetoric along the lines of that one ft op-ed warning of the dangers of "God-like AI".

What if it ends up being an exponential self-improvement fast takeoff scenario where as soon as AGI is reached, godlike maximum physically possible intelligence ASI happens basically immediately afterwards?

I subscribe to your channel because it has no noise. I just subscribed to your newsletter as well. About this video. I've given several public talks on AI in the last year. Next month I will be giving a class to seniors on fraud, with an emphasis on AI assisted fraud. I shudder at the potential effectiveness of fraud from an AGI with voice and face mimicry.

The thought of following this channel for the coming eventful years makes me happy.

Can now envision a remote worker hired who uses an AI face 100% of the time, from interview, to all interaction, before a reveal a year later!

I think we're all pretty swiftly moving from "ok, AI knows enough" to "now can it have a personality please?"

At the exponential rate of improvement if this technology, video evidence of crimes will soon be inadmissible in court because we will have doubts about its authenticity if it’s not watermarked in some way. A nice convenient way to frame someone with a crime if you can get them to talk about it, even if it never happened.

Cant wait to finally have a virtual girlfriend

Max Headroom 😮

Everything dubious...i've tested llama3 and it's unfinished to release, Boston's brainless moving metal scraps are boring (no one will give them Ai, they're a competitor to remove for every corp), Openai standing still for years, gambling on Agi becoming a meme. Chatbot as the mechanic impossible for any interaction unyil real-time processing without waiting.