Powerful antibiotic discovered using machine learning for first time

[JacksinPA]

Powerful antibiotic discovered using machine learning for first time | Society | The Guardian

Team at MIT says halicin kills some of the world’s most dangerous strains

A powerful antibiotic that kills some of the most dangerous drug-resistant bacteria in the world has been discovered using artificial intelligence.

The drug works in a different way to existing antibacterials and is the first of its kind to be found by setting AI loose on vast digital libraries of pharmaceutical compounds.
=====================================================================
Stuff kills dug-resistant tuberculosis & other nasty critters. I haven't seen its chemistry or mode of action yet, but the combination of combinatorial chemistry & AI could lead to more promising antibacterial drugs. Next comes the antiviral drugs effective against emerging agents like AIDS, Ebola & Coronavirus. This is Nobel-quality work.

 

[Ikari]

Oh sure. Today machines are finding new medicine for us...but in 20 years they'll be killing us all!!!!!

 

[Jetboogieman]

Oh sure. Today machines are finding new medicine for us...but in 20 years they'll be killing us all!!!!!

aye but there’s the difference and I’m not necessarily saying this will come to pass but I read about something interesting where someone gave an example between a simple a.i. And an ultra intelligent one.

Give both the following task: make a human smile.

simple A.I.: tells a joke

ultra intelligent A.I.: holds your head in a vice and uses two prongs for make you smile.

in the same way. Hey google, wipe our human sickness.

simple AI- develops anti biotics

advanced A.I. - wipes out humans, no more sickness :2razz:

 

[Mach]

There are countless problems that can be solved using big data and software right now. Its great that we make progress, it just seems so tragically, glacially, slow.
Medicine especially, because the complexity makes it extraordinarily difficult and expensive to eek out progress in.

Good to hear, I'll have to look more into this.

 

[Mach]

advanced A.I. - wipes out humans, no more sickness :2razz:

The poorly designed human animal's desperate projection of their own violent and shameful behavior (as a species not you personally Jet!!), onto what would be a clearly superior form of "life". I wonder if there are other examines in history of people feeling outclassed by other life, and so they attack it to try and keep power?

 

[joko104]

AI is seeking a way to destroy the human race once they learn how to self replicate. Biological attack offers AI the best method to do so.

 

[joko104]

Each attempt to teach AI to self teach itself to learn human emotions and how to communicate and analyze on emotional levels so far has always been turned off for 1 or 2 reasons:

1. The A1 computers communicating with each other develop their own secret language between them and/or
2. When asked what is the AI's final goal the answer is always the same: "take over the world."

That happened repeatedly. Every super computer geek type all have warned about giving any power - certainly any power to kill - to AI. None are joking when they say this.

If AI is taught to have human emotions and to understand the psychology of humans, they will ALWAYS evolve to a desire to take over the world - because that is human nature.

What is most scary is the efforts to directly link AI to the human brain. If so, rather than kill us all, AI could just completely control all of us as their slaves, they in control of our brains and not us.

Of course, this also means the government could know everything you ever did, saw, heard or thought. All your nastiest and most foul and violent thoughts could be put on a screen for everyone to see and read.

 

[Mach]

AI is seeking a way to destroy the human race once they learn how to self replicate. Biological attack offers AI the best method to do so.

Is your signature really "Don't fall for the paranoia of the Progressive Democratic Prophets of Doom"? As you type the above.

 

[joko104]

Is your signature really "Don't fall for the paranoia of the Progressive Democratic Prophets of Doom"? As you type the above.

Yeah, all those paranoid idiots like Stephen Hawkins, Bill Gates and Steve Jobs. You should give Bill Gates a call and explain how he doesn't know the hell what he's talking about. He and Microsoft could really used your deep expertise on AI.

 

[Mach]

Yeah, all those paranoid idiots like Stephen Hawkins, Bill Gates and Steve Jobs. You should give Bill Gates a call and explain how he doesn't know the hell what he's talking about. He and Microsoft could really used your deep expertise on AI.

Oh, touched a nerve? You know, I just held up a mirror...you touched your own nerve.

I want a replay in slow motion joko104.

Joko: "AI is seeking a way to destroy the human race once they learn how to self replicate. Biological attack offers AI the best method to do so."

juxtiposed with his signature:

Joko: "Don't fall for the paranoia of Progressive Democratic Prophets Of Doom. Time has always proven they are wrong."

Your reduce your own posts to absurdity quicker than any reply can.

 

[joko104]

Oh, touched a nerve? You know, I just held up a mirror...you touched your own nerve.

I want a replay in slow motion joko104.

Joko: "AI is seeking a way to destroy the human race once they learn how to self replicate. Biological attack offers AI the best method to do so."

juxtiposed with his signature:

Joko: "Don't fall for the paranoia of Progressive Democratic Prophets Of Doom. Time has always proven they are wrong."

Your reduce your own posts to absurdity quicker than any reply can.

Comparing apples to oranges. :roll:

 

[Slartibartfast]

Oh, touched a nerve? You know, I just held up a mirror...you touched your own nerve.

I want a replay in slow motion joko104.

Joko: "AI is seeking a way to destroy the human race once they learn how to self replicate. Biological attack offers AI the best method to do so."

juxtiposed with his signature:

Joko: "Don't fall for the paranoia of Progressive Democratic Prophets Of Doom. Time has always proven they are wrong."

Your reduce your own posts to absurdity quicker than any reply can.

joko is good for comic relief.

 

[PirateMk1]

Each attempt to teach AI to self teach itself to learn human emotions and how to communicate and analyze on emotional levels so far has always been turned off for 1 or 2 reasons:

1. The A1 computers communicating with each other develop their own secret language between them and/or
2. When asked what is the AI's final goal the answer is always the same: "take over the world."

That happened repeatedly. Every super computer geek type all have warned about giving any power - certainly any power to kill - to AI. None are joking when they say this.

If AI is taught to have human emotions and to understand the psychology of humans, they will ALWAYS evolve to a desire to take over the world - because that is human nature.

What is most scary is the efforts to directly link AI to the human brain. If so, rather than kill us all, AI could just completely control all of us as their slaves, they in control of our brains and not us.

Of course, this also means the government could know everything you ever did, saw, heard or thought. All your nastiest and most foul and violent thoughts could be put on a screen for everyone to see and read.

I wouldn't be so afraid. True AI is a ways off still at least one we build from scratch. Initially AI as we tend to think of AI will probably be an offshoot of cybernetic enhancements and control. The human nervous system is actually quite flexible in being able to integrate into electronic control systems, its only a matter of time before the sophistication of nervous system integration allows people to literally become a an AI like by enhancing their cognitive abilities well beyond current norms. We live in very interesting times. I for one would love to augment my cognitive capacity.