Has to do a lot with dopamine time horizon as well. Sometimes it takes weeks to months before even getting a glimpse of the result of the effort put in (which is already a very short time horizon compared to other industries).
While with games there's always a lot of instant dopamine even if chasing an "in-game goal" that's far away.
This is core feature of LLM
Since its a token predictor, to perform well in complex tasks it needs to encapsulate complex information in fewer words and more specialized terms
Taking something complex and putting it in easy to understand english is really hard when for a machine that can’t “visualize” stuff and relies purely on language.
Very interesting to see US vs China AI discourse
After all the biggest fear most people have with AI has always been that the most powerful models will be too capable and at the same time in-accessible to most people. Creating the fear of a permanent underclass for those without this access.
With Mythos pre-release this is specifically what Anthropic did, few had access and most didn’t. Then comes the heavily guarded and filtered Fable 5 available to the public that basically says - “I ain’t gonna do that for you.”
Off-course the Chinese AI labs now are basically perceived as the Robinhood for the masses.
American AI companies have now (rightfully) lost the moral high ground.
OI can also be faked. If you want truly measure if volume is being faked or not on an exchange - you should look the variance/volatility in OI and how often it changes
So many scam exchanges where OI would just sit at say 5M for days but the volume will be few 100M daily, or sometimes randomly go to 20M - stay there flat for a couple days, and go back to 5M
They just keep getting creative in faking numbers, wash trading in different ways to make it look real - but this one is the hardest to fake.
Its also true for some “trusted” CEXes when trying to gain traction on new products
A subtle realization here is that for the first time in history we're about to see the effects of "Intelligence in-equality"
The smarter AI is getting more expensive, using more compute and the longer it runs the better job it does - likely creating more value allowing for further use.
An exponentially upward path for the few who can, and exponentially downard (becoming useless in society) for most who can't.
Imagine a model/configuration that cost $10K per run and can likely solve 1 out of 1000 big problems. If each one of those problems, when solved, can generate $1B in value then you just need to be rich enough to throw $10M of compute to create $1B and 100x ur return. But if you're the average person this is beyond reach.
This phenomenon of wealth accumulation of "rich getting richer because of simply being rich" has always existed, but with AI the timescale in which this happens is gonna shrink exponentially causing this inequality to reach peak very soon.
SITUATION DETECTED: Google DeepMind’s AI agent autonomously solved 9 of 353 open Erdos problems in mathematics, at a cost of a few hundred dollars per problem.
IMO we’re maybe just a few decades away from solving aging at max - at-least for those who can afford it.
Be prepared for the levels of wealth/power inequality this will bring.
We our about to enter the era of the “Forever Kings”.
The year is 2532, Putin is still the President of Russia, Xi still in China, Elon Musk controls all global (Interstellar) Manufacturing while each one of his kids is a “Head of State” of some planet. The U.S. President is either some 400 year old member of the Kennedy/Trump/Obama (or similar) family or (more likely) the constitution has been amended for longer terms.
Same reasons why the aging banking industry still thrives due to regulations…...make it incredibly expensive for an underdog to disrupt the industry and increase competition
Muted accounts on twitter talking about chinese-american winter olympic drama and I think I have successfully removed 90% of elon-bucks / rage-bait spam from my twittter feed.
Pam Bondi LOSES HER MIND when asked why she has not indicted any of Jeffrey Epstein’s clients.
“The Dow is over 50k right now, the Nasdaq smashing records, that’s what we should be talking about!!”
@bryan_johnson > $50 Trillion a year generated from Human fingers imparting tiny bursts of energy into keystrokes and mouse clicks across the globe.
Can turn into $1,000T+ if they press the right keys in the right order
East Asia speedran wealth inequality to the point where if you truly innovate as an F&B business and finally drive some profits - congrats, you just increased your rent and made only your landlord wealthier
Most ride hailing vehicles in HK are EVs - and every time one has to learn a new way of operating the door handle. Please bring back simple door handles….this is pure insanity
@cloneofsimo - The value "I" get in return -
There's your answer, its not going to you. 76% is mostly going to old people (yes interest also because guess who owns the bonds)
DNS hijacking is close to impossible these days because browsers verify HTTPS certificates, and reject if the SSL cert doesn't match
They can't generate a proper fake SSL cert without compromising a CA (certificate authority)
HTTPS is extremely safe because it's encrypted even on untrusted WiFi
Worst that can happen is NSA has a backdoor in the encryption standards and the US gov can read what you do (probably)