In other words it creates work. In other words Jevons paradox.
I can’t wait for the first court case where an LLM surfaces a vuln, lazy devs ignore it, and someone later sues the company into oblivion for liability.
People 50 years ago would laugh at the irony of that statement, being made here on what we call the internet. Real life will never be digital, and yet here we are, talking to anonymous strangers many of whom are not actual people.
I considered this too but I didn't think it too ironic considering anonymous pen pals have been a thing since the 20th century at least. Obviously the technology would amaze, but the concept would be understandable and appreciated.
The abstract you linked, pasted below, seems to say otherwise. Placebo groups report same SE frequency.
“Across 684 randomized controlled trials, reported SEs were infrequent. Although dose and duration tertiles were statistically associated with study-level side effect reporting, the effect sizes were uniformly small, events were infrequent, and the reported symptoms were primarily mild and nonspecific. No consistent exposure–response pattern indicative of clinically meaningful risk was observed. Adjusted logistic regression and frequency-based analyses showed no consistent dose- or duration-dependent increase in SE risk, with placebo groups often reporting similar or greater SE frequencies at the study-reporting level. CrM appears to be well-tolerated and, at the study-level, does not increase the risk of gastrointestinal, renal, liver, musculoskeletal, or other SEs compared to placebo, even at high doses or longer durations.”
1) Creatine is safe, "creatine supplementation is safe across a range of doses, durations, and populations according to human trials".
2) Stomach issues are a commonly reported side effect, "Gastrointestinal issues were the most frequently reported side effect... Most notably, gastrointestinal distress is a commonly reported side effect, and those supplementing with creatine may need to divide the dose into smaller boluses to alleviate it; however, it is worth noting that this side effect was not persistent"
3) It is not clear if the side effects are caused by creatine, many of the issues can be explained by baseline issues rather than caused by creatine, "No consistent or clinically meaningful dose-dependent increases in side-effect reporting were observed across models; even at higher doses and prolonged durations, reporting remained low and largely comparable to placebo at the study level"
Cognitive dissonance can explain a lot. If you don’t think the current regime is genocidal (whatever that even means) then you might get very concerned that anybody who says it is genocidal is a dangerous lunatic or terrorist sympathizer. Even saying something obviously truthful like “there are good people on both sides” becomes a threatening provocation. Hate is a system.
To be clear, it was God that decided to give this land to Abraham in "everlasting possession," so this is pretty cut and dried. Why would Abraham lie about that?? /s
Hard for to assess that stat given the polling organization has this on their Wikipedia page.
“On 29 August 2024 the Israel Defense Forces released Hamas documents[6] that it said showed that, unbeknownst to the PCPSR, Hamas had secretly falsified its levels of public support in polls conducted by the PCPSR.[7] Rejecting the IDF claims, Shikaki said it was 'highly unlikely' that Hamas had falsified its results, but vowed to probe the claims.[8]”
Israel gov./IDF/etc. is known to falsify Hamas documents, force false confessions from UN workers using brutal violence and to regularly lie to justify its actions or shape public opinion. Anything they say about the other side means very little without actual independent investigation/corroboration or direct observable evidence. Independent journalists are allowed to go to Israel, so there's no need to lower the standards.
I read all these PCPSR polls in detail as they were released. And if you actually read the polls you'd never be able to conclude what you did, for one simple reason. Those polls consistently show vaaaast majority (>90%) of Palestinians from Gaza did not believe Hamas perpetrated attrocities attributed to them by Israelis. There was never a direct qustion about support of killing Jews in those polls either.
OTOH, extermination of population of an enemy city was a direct question in that Israeli poll.
Also on another note, a poll of a population of Gaza does not represent all Muslims. You can easily say that a long term occupied and terrorized population by Israeli Jews, will not be best representative sample of Muslim views on Israel or Jews, even if that population is very largely Muslim.
Conflating views of all Muslims with Muslims from Gaza, lying about the poll and misrepresenting the results to fit your whataboutist views, etc. only shows your total lack of rigor. Nothing else.
It’s always the poor and uneducated voters that get the blame. It’s never the actual billionaires who got Trump elected and who control everything he does and much of the media zeitgeist because, you know, 100s of billions can really flood the zone. I don’t think you’re evil for being fooled, but try to think a bit deeper about where power and blame lie. It’s not uneducated rural folk the median yuppie finds uncool.
Why not both? I read an article recently about the Texas Senate race and one Republican voter they interviewed said it was about "the immigrants and the guns." So low information voters get a pass because they're awash in right wing propaganda? What happened to the oft cited right wing value of "personal responsibility"?
I totally get it. What I don’t get is why these perspectives are limited to writing code and not writing or communicating more generally. Just like coding, Eric Schmidt himself is coming to an end. How can smart people not see, writing code is just a canary in the coal mine, a special case that’s only a difference in degree not kind from other sorts of knowledge work?
Because the people who funded them from the outset did so because that was their goal? The destroy all jobs talk started at least 10 years ago, it many ways that hype is the actual product itself.
I can’t wait for the first court case where an LLM surfaces a vuln, lazy devs ignore it, and someone later sues the company into oblivion for liability.
reply