But it’s a good question. I want to know.
I am assuming this is not possible. The only thing i know of is capable of doing so is pegasus. But it’s very expensive afak.
the main thing about LLM's in my opinion is the tokenization part, words are already clustered and converted into numbers(vectors) it's already a big deal. we are using learned weights,
the attention part feels like a brute force approach to learn how those vectors are likely used together (if you add positional encoding as an additional information).
statistics on large amount of amount of data just seems to work after all.
Yes true. I think they are biased towards the technical complexity of designing such systems. Although to experts in the field it is just a bunch of statistics and some math and coding. It’s not really a big deal but they miss the point that this can be dangerous and life impacting and this is depressing
By complexity, i meant the hardware it self. Not even the architecture or the instructions set. For example you can design a virtual machine in days. But making a real one is at the core of the geopolitical issues we have today
This proves that at the core. The things we rely on to achieve great software and life impacting technologies are extremely simple. The complexity is that how to make them.
Always keep in mind that you can’t go faster than what the harware can allow. Sometimes even by what the OS can allow. We can do way better on some specific tasks. But for for that you need to rewrite some algo implementation to better take advantage of all the hardware can allow. ML is a specific task, if you can take advantage of the harware to do better matrix mul. Does not mean you can have the same speed ups overall. Mojo for example
Since most models require little code compared to big software projects, why not use c++ or any other compiled language directly.
Python with it’s magic functions, shortcuts is just hiding too much complexity which can result in bug performance issues. Plus code is more hard to maintain
Everything is perfect even better than real life. But still humans are still far from being looking natural, they always look like a balloon to me. Something is always off when it comes to simulating humans, the eyes movements, body
I might be totally wrong, but i think that People with social skills will not find it worthy and somehow bizarre. But those with some sort of autism can enjoy the experience. Or even people addicted to drugs or psychedelics.
It can be a game changer for some professionals and really a good tool. But impact in kids can be dangerous as they will be confused to what is reality
What if programming languages start offering both?
In my opinion RC is GC in disguise.
At least for example Golang GC has the merit to run in a separate thread(still has to stop the world when reclaiming memory back, and the memory allocator model is helping achieve great GC perfs).
Influencers culture has killed social media platforms, most ppl want to be influencers today to make easy money.
Back in the days social media platforms were used to connect with friends and maybe strangers just for the sake of socializing. Not anymore
It’s always good for this kind of articles, to refer to source code. Provide links to the source code responsible of some of the implementation details
I am an energetic person. I really find it very hard to fall asleep. But once i do, i really go into deep sleep, 7 to 8 hours of deep sleep.
Also i am in my thirties. And i am still very fit and athletic body shape.
Ppl are always surprised to know know that i am 30+
I only eat healthy food. Because i am very selective when it come to food and i don’t feel hungry much.
My brother used to be in a great shape. But past 30 he become overweight.
Sleep is good but what you eat and how much you eat is the key
Well, in my case. I realized that i find my day to day job boring and there is still little joy. Mainly because i don’t feel challenged. I know how to do it so what i see is that how much effort is needed to accomplish.
What i do is that i invest my free time/ no commitments, in things harder than my current skills allow me to achieve. And thankfully i do find joy in doing that. In my case is just joy not making something big and successful
Treat work as normal labor. Don’t expect joy from work.
Try to be less involved. But still efficient, just the bare minimum. It’s just a contract after all.
Invest the money you get from your labor in a stable social life. Healthy life style.
Spend the free time you’re left with practicing your hobbies. It can still be coding. But something you’re excited about.
I’ve developed a phobia from street dogs. Especially with the bad experiences i had with them when i was kid.
But what i have noticed. When the street is busy. They’re peaceful otherwise they will try to intimidate you as a group either you’re walking,cycling, biking, driving.
When i am in the car and i am safe. I play games with them. I stop the car. They stop and they wait for me to drive. So they can run and bark outloud. Seems like they’re just playing. I wish i had the same courage when i am walking/running
I created a small program using arduino and IR receiver to control PowerPoint slides using any available remote control.
I had to create a small go program to interface with arduino-cli so that I don’t need Bluetooth or any other hardware, just the serial port
I personally use apple keynotes to produce videos.
It’s a powerful tool, and just yesterday i was wondering why imovie was lacking so much features.
And that is what i like about apple, they target consumers, not professionals
Technically, all programming languages are constrained by what a machine can do (the cpu architecture), to me all programming languages are transpilers, just chose the language that fit best your usecase, and that will allow you to write correct code easly, but always keep in mind that regardless of the language you’re using, you’re still constrained by same machine architecture,