The fallout from "No, not Hank Green using that AI" is at the stage where people are debating if "canceling" people is bad.
-
@futurebird It also has a meaning in infosec, when working with indicators of compromise. AFAIUI, anyway. I think in the sense that you might "pivot to effective altruism" to learn more about HG. If he was a threat actor.
(disclaimer: I'm not an infosec guy, but I've done a bit of reading in related areas)
The worst thing about Pivot is it is some deep tech people inside baseball stuff an only the people who know know.
That might need to change.
-
Oh! If you don't know what "Pivot" is? Let me tell you about it: The name is a play on the phrase "Pivot to video" from the famous attempt to shut down print media and increase ad revenue. It's short hand for a bait and switch.
Pivot is basically the local news around here for tech topics and markets. David is a cool dude who is bad at interior decorating and he just interviewed Molly White.
@futurebird I adore Molly White.
-
@futurebird I adore Molly White.
She is such a super star AND a real one.
-
Oh! If you don't know what "Pivot" is? Let me tell you about it: The name is a play on the phrase "Pivot to video" from the famous attempt to shut down print media and increase ad revenue. It's short hand for a bait and switch.
Pivot is basically the local news around here for tech topics and markets. David is a cool dude who is bad at interior decorating and he just interviewed Molly White.
Claude Opus and I are going to write up a neural net for AI Bullshit Detection
Most of the Pro- camp don't understand how a Transformer works and absolutely none of the Anti- camp get it, either.
-
@futurebird It's been quasi-religious (at least) for a long time now, so of course it has the same trappings of Christians having differences of approach whether they focus on God as loving father, power-justified tyrant, or some combination of the two, but recreating the Predator handshake meme in their agreement that God is omnipotent.
All the more reason it's clearly nonsense (which is terrifying to see people fall for) for anyone outside the reality distortion bubble where emotions motivate belief in some kind of Turbo Pascal's Wager.
@MaddieM4 @futurebird now I'm thinking about some betting software written in this
-
Claude Opus and I are going to write up a neural net for AI Bullshit Detection
Most of the Pro- camp don't understand how a Transformer works and absolutely none of the Anti- camp get it, either.
... pecking away:
"
A few candidate sections, playing off Silicon Valley's specific dialect of bullshit rather than generic corporate-speak:1. Detection Targets (Threat Model)
The Visionary — speaks exclusively in the future perfect tense ("by 2027 this WILL HAVE transformed X"), never the present, never a mechanism
The Thought Leader — a LinkedIn-native subspecies; detectable by unprompted use of the word "unlock"
The Doomsayer — Chicken Little's tech-adjacent cousin; identical confidence profile to the Visionary, opposite valence
The Stealth-Mode Founder — bullshit so early-stage it hasn't even shipped a lie yet, just vibes and a Figma
2. Signal Requirements
MUST flag any sentence containing "democratize" applied to something that costs $200/seat/month
MUST flag "we're not just building a product, we're building a movement"
SHOULD down-weight claims inversely proportional to Series-letter (Series A claims get more benefit of the doubt than Series F claims, which SHALL be treated as confirmed bullshit by regulatory default)
3. False Positive Exclusions
Grace Hopper is exempt. COBOL manuals are exempt. Anything that shipped and is still running unglamorously twenty years later is, by definition, not bullshit — regardless of how it talks about itself
4. Non-Functional Requirements
Latency: the detector MUST fire faster than the bullshit can secure its next funding round
Uptime: 100%, because the bullshit never sleeps either"
-
@exchgr @futurebird If I understand his argument correctly, what he's saying is that AI *could* satisfy everyone's needs and provide universal abundance (which, no, it can't), but that we should avoid that if at all possible, because that would deprive everyone else of agency. So out of concern for everyone else, he favors the future where he keeps all of that wealth and abundance for himself.
Presumably, that's the "effective" part of effective altruism.
-
I legit think the doomers and boosters come from the same source at this point, just with different flavors of the same message that may resonate with different groups of people.
You won't buy that AI will save the world and grant wishes? Maybe you'll go for "it's so dangerous" ... and therefore still must be very powerful and effective.
@futurebird yes, neither scenario will come true. It will be different this time, though. Rather than just one third as large as the entire economy, the AI bubble companies are 10 (TEN) times bigger. The crash is going to be spectacular. Compared to 2008? Hahaha, you haven’t seen nothing yet. It took eight years for the nasdaq to recover. Ten times? Not in your lifetime…
-
The fallout from "No, not Hank Green using that AI" is at the stage where people are debating if "canceling" people is bad.
This because about 10k people unsubscribed from his youTube channel, something that probably caused him to take stock and leave the internet for a bit.
Unsubscribing from a youTube channel isn't "canceling" someone. Can we all calm down please?
Also, that he used AI isn't really the big issue, it has just formed a concise talking point around something bigger.
@futurebird I have never heard of Hank Green, but I am not going to worry about AI until ants start using it.
-
@futurebird It’s as if some folk feel they are owed an audience. Folk will support folk they want to support.
Carolyn It’s weirder than that. I don’t think that Hank Green has any issue with people unsubscribing from his channel over this – it’s A) a relatively small number for him, and B) IIRC is the kind of behaviour he himself has suggested in the past is the way to handle shit like this.
What’s happening is that people who like Hank Green are now wondering if cancelling is bad, as if “cancelling” has ever been something other than a mix of tuning out and obsessive harassment. They don’t want their precious little boy to lose views, because they think he’s entitled to them, and they don’t want to see the weird obsessives show up in their little comment playground.
They’ll be OK with all of this the next time someone they’re not parasocially invested in does something they don’t like.
-
@futurebird It's been quasi-religious (at least) for a long time now, so of course it has the same trappings of Christians having differences of approach whether they focus on God as loving father, power-justified tyrant, or some combination of the two, but recreating the Predator handshake meme in their agreement that God is omnipotent.
All the more reason it's clearly nonsense (which is terrifying to see people fall for) for anyone outside the reality distortion bubble where emotions motivate belief in some kind of Turbo Pascal's Wager.
-
The fallout from "No, not Hank Green using that AI" is at the stage where people are debating if "canceling" people is bad.
This because about 10k people unsubscribed from his youTube channel, something that probably caused him to take stock and leave the internet for a bit.
Unsubscribing from a youTube channel isn't "canceling" someone. Can we all calm down please?
Also, that he used AI isn't really the big issue, it has just formed a concise talking point around something bigger.
Oh no, please tell me Hank isn’t into “effective altruism”.
That would be far more disturbing to me than Hank being into “ai”. A lot of smart people are still confused about LLMs and it’s not a moral failing.
But since “effective altruism” is basically the underlying cult of tech-fascism built on dubious interpretations of the already dubious utilitarianism - it’s a lot more disturbing to me.
-
@exchgr @futurebird If I understand his argument correctly, what he's saying is that AI *could* satisfy everyone's needs and provide universal abundance (which, no, it can't), but that we should avoid that if at all possible, because that would deprive everyone else of agency. So out of concern for everyone else, he favors the future where he keeps all of that wealth and abundance for himself.
Presumably, that's the "effective" part of effective altruism.
@lrhodes @futurebird it’s effectively altruistic!!!
yeah, that movement’s whole thing is “get as rich as you can so you can donate the most money to the most effective causes,” but i’m pretty sure everyone in it forgot the last part
-
Carolyn It’s weirder than that. I don’t think that Hank Green has any issue with people unsubscribing from his channel over this – it’s A) a relatively small number for him, and B) IIRC is the kind of behaviour he himself has suggested in the past is the way to handle shit like this.
What’s happening is that people who like Hank Green are now wondering if cancelling is bad, as if “cancelling” has ever been something other than a mix of tuning out and obsessive harassment. They don’t want their precious little boy to lose views, because they think he’s entitled to them, and they don’t want to see the weird obsessives show up in their little comment playground.
They’ll be OK with all of this the next time someone they’re not parasocially invested in does something they don’t like.
@kichae Cancel, schmanzel. People are allowed to like something then change their minds. And the damage done by data centres is more pressing than someone’s views, especially when that person’s use of tech contributes to that damage.
-
Oh no, please tell me Hank isn’t into “effective altruism”.
That would be far more disturbing to me than Hank being into “ai”. A lot of smart people are still confused about LLMs and it’s not a moral failing.
But since “effective altruism” is basically the underlying cult of tech-fascism built on dubious interpretations of the already dubious utilitarianism - it’s a lot more disturbing to me.
@CptSuperlative @futurebird I never got into him for this reason. Even though he personally might have the best intentions and keeping an open mind. I feel he isnt critical enough. Hes the perfect example of person with intelligence and very little wisdom.
-
@MaddieM4 @futurebird now I'm thinking about some betting software written in this
@ehproque @MaddieM4 @futurebird You beat me to it! I still miss Turbo Pascal (and Turbo Vision - it was fast and just worked).
-
The fallout from "No, not Hank Green using that AI" is at the stage where people are debating if "canceling" people is bad.
This because about 10k people unsubscribed from his youTube channel, something that probably caused him to take stock and leave the internet for a bit.
Unsubscribing from a youTube channel isn't "canceling" someone. Can we all calm down please?
Also, that he used AI isn't really the big issue, it has just formed a concise talking point around something bigger.
@futurebird what's wrong with being a doomer on AI? I've been an AI skeptic for the longest time, but since I've still been doing transformer research in the meantime, I've seen it progress from irrelevance to being on the cusp of recursive self improvement. It doesn't matter it's now not fully capable of RSI, but the ratchet effect of RSI means it's too late to worry about it after it's been achieved
-
@futurebird what's wrong with being a doomer on AI? I've been an AI skeptic for the longest time, but since I've still been doing transformer research in the meantime, I've seen it progress from irrelevance to being on the cusp of recursive self improvement. It doesn't matter it's now not fully capable of RSI, but the ratchet effect of RSI means it's too late to worry about it after it's been achieved
Why do you think RSI will happen?
-
@futurebird what's wrong with being a doomer on AI? I've been an AI skeptic for the longest time, but since I've still been doing transformer research in the meantime, I've seen it progress from irrelevance to being on the cusp of recursive self improvement. It doesn't matter it's now not fully capable of RSI, but the ratchet effect of RSI means it's too late to worry about it after it's been achieved
Forgive me for being a little cynical about "and then RSI will happen" but it's not like this hasn't been tried before. In robotics there was a lot of excitement about "let's just make a robot then have it evolve a way to move the motors to do the task" this could work in very limited demo settings, but to really solve a problem it's just a big mess, this kind of random variation problem solving is just a lot slower than just writing it yourself.
-
Why do you think RSI will happen?
@futurebird well, LLMs can now upgrade and improve legacy architectures with ease.
In 2019, training GPT-2 cost $43,000 and took a couple of days, today, Karpathy improved it to <$100 and 2.02 hours of training. "Autoresearch", a genetic algorithm that used LLMs as mutators, found another ~11% improvement to 1.80 hours.
Currently, the record is 73.8s to achieve 2019-level AI frontier performance (https://xcancel.com/classiclarryd/status/2086582390135406713)
