Sabitlenmiş Tweet
aut0
7.2K posts
aut0
@aut0m8d
I can remain irrational longer than they can remain solvent
the zone Katılım Şubat 2010
240 Takip Edilen3.8K Takipçiler
I have a storage unit at a place that has two access gates that both look the same. the main one slides, and the other one opens like this
the first time I used the second gate it caught me so off guard I laughed out loud
Charlie Gribble@Charlie_Grib
Former NC State head coach Will Wade departing from Raleigh for the final time on his way down to Baton Rogue.
English
this is hilarious. dude gives his $1.5M bonus to his employees.....at 1,500 stores, averaging 25 employees per location
$1,500,000 / 1,500 / 25 = $40
bro gave everyone $40 and DID PRESS ABOUT IT. genuinely thinks he's being generous. boomers are unreal
nbcboston.com/news/national-…
English
@fchollet I honestly don't think the criticism comes from people who actually understand your intent, I think it comes from people who have unspecific definitions of AGI and believe you're handwaving in the same ways they are. I wouldn't take it seriously, you're doing great work
English
If you care about the rate of AGI progress, you should be excited about a new eval that focuses research efforts by pointing out important gaps & providing a way to measure progress towards fixing them
If instead you only care about having your preconceptions confirmed, too bad
English
just put $1000 on this
you do not fuck with FWC
MyFWC@MyFWC
The FWC is aware of a video depicting individuals in the Everglades on an airboat who appear to be discharging firearms at an alligator. FWC officers are looking into the incident and will provide additional information when available. To report wildlife violations, call the Wildlife Alert Hotline at 888-404-3922.
English
I really hope these mechanics work. the amazing thing about Twitter/X was the human connection, and we really need it back
ApoStructura@ApoStructura
Trying this new “Accounts you follow and who they follow” reply filter on my latest post that Elon reposted. As soon as I turned it on the number of spam replies collapsed. I don’t think I’ll turn it on for all my posts, but for posts like this it looks like a great feature.
English
the other cool thing about making proactive scans more readily available is that I could easily provide my validated data to people studying Alzheimer's, and they could easily take advantage of it
only 1-2% of humans have two copies of APoE4 so it's difficult to study atm
aut0@aut0m8d
I have two copies of the APoE4 gene which means I have a ~60% chance of getting Alzheimer's by 85, so it would be nice to do these regularly as I age would be awesome to see this come down in price for that reason
English
labcorp has been doing this for blood tests for a while, it's neat bc you can see trend lines if you do many tests: ondemand.labcorp.com
Also Function is doing this for MRI/CT scans as well: functionhealth.com
gabriel@gabriel1
on top of normal healthcare, for low stakes curiosity, i'd love a cheap clinic with all screening equipment. no doctors and no diagnoses i can book any screening, and they send files that ai reads. and i add it to my health context so i can continue asking chatgpt questions
English
on top of normal healthcare, for low stakes curiosity, i'd love a cheap clinic with all screening equipment. no doctors and no diagnoses
i can book any screening, and they send files that ai reads. and i add it to my health context so i can continue asking chatgpt questions
English
it's really annoying even in a short thread. if you tell codex "okay this plan looks good, just don't do this one thing", it'll spend 40 minutes researching that one thing and remake the plan around building a system that never does that thing under any circumstances
Andrej Karpathy@karpathy
One common issue with personalization in all LLMs is how distracting memory seems to be for the models. A single question from 2 months ago about some topic can keep coming up as some kind of a deep interest of mine with undue mentions in perpetuity. Some kind of trying too hard.
English
the incentive fix here is really important and correct. people who do not live in a place should not have an incentive to get impressions in local discussions because they have no dog in the fight. engagement via creating controversy and disagreement is all upside for them
Ramp Capital@RampCapitalLLC
@tunguz I think the intent is to stop monetizing foreign actors for creating chaos in the U.S.
English
I'm based in France, but 43% of my audience is American.
I know many of us are in the same situation. To reach a wider audience, all my posts are in English, the international language.
Those who cause trouble may be punished, but with this change, which will significantly reduce our earnings, you're also penalizing a number of accounts that use the international language without any ill intent.
X tells us that videos and new, high-quality content will be prioritized, but this change will, on the contrary, drastically reduce our earnings.
Is there any way to reverse this decision, please?🥺
@XCreators @X @nikitabier @elonmusk
Nikita Bier@nikitabier
Starting Thursday, we'll be updating our revenue sharing incentives to better reward the content we want on X: We will be giving more weight to impressions from your home region—to encourage content that resonates with people in your country, in neighboring countries and people who speak your language. While we appreciate everyone's opinion on American politics, we hope this will disincentivize gaming the attention of US or Japanese accounts and instead, drive diverse conversations on the platform. We invite creators to start building an audience locally. X will be a much richer community when there's relevant posts for people in all parts of the world.
English
my conviction in this thesis has only grown. added a ton in the last 2 weeks
aut0@aut0m8d
the short-term mindset of most crypto traders is an absolute gift to those who are in the know and thinking long term. like $HNT has absolutely no business being as cheap as it is. it is an obviously good deal. but aut0, if it's an obviously good deal, why isn't everyone bidding? it's this fantastic information arbitrage situation. almost no one with tradfi institutional money understands these smaller cap projects, because it takes a lot of reading to understand them, and the content isn't all in one place, and liquidity is fairly thin anyway - they can't really ape with size (10-100M). so the people moving the charts are degens with big bags. the few intelligent VCs in the space are already positioned and have been for some time. being a great deal at these prices doesn't mean it makes sense for them to overallocate so the set of degens with bags who could move these charts are generally people who have been here for a minute, and have heard of these projects, and have probably even bid on them in the past. a lot of them have even *lost money*. maybe even multiple times! and the thing with crypto is that most of the things are bullshit, and when you lose money, you're inclined to dismiss the whole project as bullshit - hard to determine if the team is selling on you, and losing hurts so the quick analysis is usually the most negative - "I got robbed!" even if you work out the technical reasons for why you lost out on a trade, the emotions still remain. it's hard to say to yourself "I got rekt on this before, but now I'm going to get back in". a lot of great traders in this space will refuse to re-enter a trade they got out of bc the emotions fuck up your objectivity. you end up chasing losses a little. it's messy HNT had a number of challenges with their early models. first they started with IOT and though they successfully rolled out the network, demand never came - there were simply (almost) no customers. as a result of this, token emissions (rewards for radio deployers) WAY outweighed token burns from data usage, and you ended up with a basically constantly inflating circulating supply of tokens with little to no relative buy pressure. extremely bad for price action but Helium didn't give up at that point, they pivoted the whole model. they realized the money wasn't in IOT, but in 5G, and started to work out the challenge of executing on a model with much much different economics and technical challenges. it took years and many failed experiments to figure out how to do it. HNT was split into two additional tokens, the model got super complicated, and a lot of investors lost money as the circulating supply continued to inflate and radio operators continued to sell for profit. it was messy, but completely unprecedented ideas are usually like that at first skip foward to now, and they've seemingly cracked it. the token was simplified back to one single HNT. the docs are like 3 total pages, and written clearly enough for a layman to follow. the emissions (which halve every 2 years) have fallen underneath burns from data usage. real-world usage of 5G data has been up and to the right, and growth-rate is starting to accelerate. they've identified niches they are uniquely positioned to capture (i.e. wifi deployments at airports) and have built tooling to help people onboard those systems. deployers are competing for ideal real-estate to capture more data usage, and the 1.5-2 MILLION *daily* users connecting to the network are informing the best places to do just that (with anonymized location data). the flywheel is starting to spin up, and it's clear as day to anyone paying attention but the list of people interested in buying low-cap cryptocurrencies remains tiny. sentiment around crypto among the general public is basically at an all time low. it's really just those degens, who have been burnt in the past, who are chosing not to bid because they genuinely think this variety of project is fundamentally flawed and is not coming back. this will change as the numbers grow, because the economics of the project will start forcing the price up enough that it becomes undeniably interesting again HNT flipping deflationary is basically the alarm sounding for the beginning of that process. it's happening right now, but almost everyone on CT is still sidelined because they haven't checked in on the latest progress. they should check in x.com/blknoiz06/stat…
English
this is fundamentally why parables have defined human culture for ~all of human history. the secret techniques fall out of the stories and they often cannot be understood without them
𝚟𝚒𝚎 ⟢@viemccoy
There are secret esoteric techniques hidden in the corners of the deep web which will teach you how to do things you've only dreamed of, the catch is that they only work once you've become the type of person who could have invented them without the PDF in the first place.
English
@ryanzip I think personality is gonna be the next big frontier tbh
x.com/aut0m8d/status…
aut0@aut0m8d
p sure the top labs will soon realize they need to start churning out "alt" base models that see the data in unique ways to get truly novel results
English
GPT has the personality of a company that would name their model GPT.
Claude has the personality of a company that would name their model Claude.
English