Glemllksdf
On one side you can interpolate in what direction it can go, than you can also add the general speed we are currently seeying and then you can try it out yourself.
The conclusion?
1. its clear that currently its critical to be aware of whats going on. With this you can act sooner or be part of this
2. if it hits hard but not too hard, you might have an advantage because you know how to use it
3. if its stalls you can reduce your effort in this area
If autopilot was missleading, full self driving is too?
For a VM running for a month? No.
For highly scalable batch tasks? Yes.
For small experiments? Yes.
For Full e2e tests creating a full env and killing it a minute later? yes.
And weirdly enough, a Task like sorting a file with data in it, if you are not a professional, windows offers very if not non single way of doing this. You would need to understand file types, understand that csv can be imported for excel, you need excel, than you need to understand excel how to sort stuff in it.
The ffirst thing I do in Excel is select the pseudo table and click on table -> insert to make it a sortable / real table. I showed this to every one in my Team full of studied CS people because non of them knew this.
Flow simulation also happens on GPUs and not CPUs though.
El Capitan is the top 1 on top 500 and the flops ratio between CPU and GPU is nearly 1 to 100.
The only good thing is that it hurts americans too. Alone the fertilizer is a huge cost increase.
Its not just jet fuel.
But the export/import will also be hit, all these special high tech products from and to europe
We know how to do a lot of things, how to automate etc.
A billion people do not know this and probably benefit initially a lot more.
When i did some powerpoint presentation, i browsed around and draged images from the browser to the desktop, than i draged them into powerpoint. My collegue looked at me and was bewildered how fast I did all of that.
Its clear that it will go in this type of direction but Anthropic announced managed agents just a week ago and this again with all the biuld in connections and tools will help so many non computer people to do a lot more faster and better.
I'm waiting for the open source ai ecosystem to catch up :/
CIA and FBI and states easily pay 100k for a zero day.
Plenty of companies have security expert staff on file.
And it will become cheaper and easyer to use, fast.
Having different opinions on AI/LLMs doesn't make the use of it the same as replacing flour with sawdust.
The AI 'image' slop for example, i don't think its bad. But i also don't think it takes anything from a real artist. It takes jobs from people with drawing skills but it doesn't change anything for an artist.
But LLMs compute requirement is so high that it pushes the boundaries of compute, memory and memory bandwidth which is fundamental for curing diseases.
LLMs math / neural networks can and are used for medical research. Simulating a whole body with proteins, cells etc. will bring us the breakthrough we need.
Nothing in modern medicin research is withoout compute.
AlphaFold def helps researchers around the globe.
Musks SpaceX Keynote was ridiculous, don't get me wrong, but we will be able to see AI progress in the next 5 years which will give us some kind of gut feeling were the journey can go.
Also AI solves another problem: Compute. It was clear that we want some kind of compute but its like with 4k; We have 4k for ages now but it is not the default resolution on all displays sold. We stoped pushing the boundaries because invest is not here. People do not bother too much with it.
With AI and the richest companies and people want to see what happens, pushes the envolope a lot faster, pushes us to find solutions.
This AI Compute based on ML/Neuroal Networks can also be used for physics simulation, protein folding, and everything else.
Stoping technology is not an option and not a solution. Education is. We need to educate people.
The mighty cloud provider can't solve this issue?
Google has second precision billing on compute.
Its not hard to define a base layer of allowed billing increase and adding this type of context to resource allocation.
You are not just suddenly creating a mlllion terabytes of data or a million db requests without supervision.
It could even be as simple as basic level caps like 100 euro / month, 1000, 10.000 etc.
And there is a difference between stoping everything before the spike happens vs. also deleting stuff.
Any tips around your setup running this?
I use lmstudio with default settings and prioritization instead of split.
Or is it only layers but that would affect all Experts?
A good security expert earns how much per year? And that person works 8/5.
Now you can just throw money at it.
CIA and co pay for sure more than 20k (thats what the anthropic red team stated as a cost for a complex exploit) for a zero day.
If someone builds some framework around this, you can literaly copy and paste it, throw money at it and scale it. This is not possible with a human.
And with gpt-2 the worry was mass emails a lot better and more detailed and personal, social media campaigns etc.
How many bots are deployed today on X and influencing democrazy around the globe?
Its fair to say it had an impact and LLMs still have.
Rumors say it has 10 trillion parameter vs. 1 trillion.
Balancing KV Cache and Context eating VRam super fast.
Wikipedia on IEEPA: "An Act with respect to the powers of the President in time of war or national emergency. "?
I mean thats very wishi washi. So are we both aligned that it looks like missuse? Because if its only about a word definition of no its not illegal what he did but a clear missconduct than it feels like word play.
I also think its a huge barrier allowing some LLM model access to your desktop.
Managed Agents seems like a lot more beneficial