It's a sign that we're moving on to greener pastures.
Sucks for the new generation that think everything needs an app to work.
The internet was better at feeling less corporate a decade ago.
563 karma · joined September 8, 2014
It's a sign that we're moving on to greener pastures.
Sucks for the new generation that think everything needs an app to work.
The internet was better at feeling less corporate a decade ago.
6 up 16 is a very large number.
16 up 6 is a considerably smaller number.
(I read it that way in my head since it's quicker to think without having to express "to the power of" internally)
The networks pick up on style, layout and common patterns in increasing complexity to solve the next token.
Very fascinating to watch and if we plug all the holes in an LLM model we can approach a tool that might as well be intelligent since it gets the job done due to the sheer amount of capability encoded in its weights...
True intelligence is still a work in progress however.
Maybe you're a tactile or sequential learner? There's "Method of Loci" or even "External Vocalisation".
You're capable of reading and writing so that's 100% evidence you're capable of learning - and being able to perform a sequence of tasks allows you to understand how to perform an action correctly.
Let's go with that and try external vocalisation. Many would trade the noise in their mind for silence and the ability to focus clearer on a task.
When you're working on a task, since you lack an internal monologue - vocalise it instead, it doesn't need to be audible, even moving your lips quietly should work as this will tap into your motor cortex which is highly tuned for performing sequential micro-actions.
Try talking to yourself about the steps you must take to get something done and see how you go.
---
Of course this is something you can try at home but following up with a few medical advisors will also open up other avenues for course correction if you feel this is becoming a real problem.
He's very abrasive in his conduct but don't mistake it for incompetence.
Even the "AI can't do video" thing was blown out and misquoted because discrediting people and causing controversy fuels more engagement.
He actually said something along the lines of it "not being able to do it properly" / everything he argues is valid from a scientific perspective.
The joint embeddings work he keeps professing has merit.
---
I think the real problem is that from a consumer perspective, if the model can answer all their questions it must be intelligent / from a scientist's perspective it's not capable for the set of all consumers so it's not intelligent.
So we end up with a dual perspective where both are correct due to technical miscommunication and misunderstanding.
I still "feel the AGI". I think Ben Goertzel'a recent talk on ML Street Talk was quite grounded / too much hype clouds judgement.
In all honesty, once the hype dies down, even if AGI/ASI is a thing - we're still going to be heads down back to work as usual so why not enjoy the ride?
Covid was a great eye-opener, we dream big but in reality people jump over each other for... toilet paper... gotta love that Gaussian curve of IQ right?
Imagine that current AI is already curating and generating datasets for the next generation.
Also consider that what we have now is only possible because hardware capability increased.
AGI is absolutely possible with current technology - even if it's only capable of running for a single user per-server-farm.
ASI on the other hand...
https://en.m.wikipedia.org/wiki/Integrated_information_theor...
Didn't realise this was the case.
Seems like a situation that just got worse and worse until it was a disaster without remediation.
Meanwhile SBF was quite happy to watch the world burn.
Reality is always the middle ground between the extremes.
Blissful ignorance was nice while it lasted.
The time alone to train my own local experiments is just one aspect of many.
It makes more sense to train models on clusters and perform research / data acquisition in private, then allow for inference at the edge or licensing on external infrastructure.
It seems at this point OpenAI are too big to fail.
Enjoy!
If they're running at a loss, it's a great way to take shots at the competition and especially with the added advantage of model capability.
Get more customers onboard, play around with margins as required.
You need to search for a job and do your project at the same time.
Don't neglect sleep or you will burn out.
You can do everything you want if you time-block appropriately and be consistent.
I made https://rhea.run from nothing and grew to a few hundred users with slight profit.
I can tell you right now that marketing and user onboarding will be your hardest challenge.
I found that organic growth trumped ads on reddit etc.
Facebook ads are a dead end, you will be engaging with bots or deep state conspiracy folk.
Working on product will always challenge you with unexpected hurdles, something that sounds like a day of work may end up taking a week or more in research and implementation. It's all a learning experience.
If you land a job, you can gain further experience while getting paid.
Feel free to reach out and prompt me for any further advice.