#unethicaloptimizationprinciple search results
@johnthornhillft sounds a note of caution on AI. Confabulation is probably inevitable. So is vulnerability to Adversarial attack. Nor can they easily escape the #unethicalOptimizationPrinciple. #AIEthics
#unethicalOptimizationPrinciple paper now in top 0.7% of all papers tracked by Altmetric. People need to understand: Profit maximizing AI is disproportionately likely to find unethical solutions (without v careful steps to avoid this.) royalsocietypublishing.org/doi/10.1098/rs…
Just heard from a Nobel Laureate that he's submitted an article which contains hyperlinks to @DameWendyDBE and to our #unethicalOptimizationPrinciple paper. Who I wonder will write the 2nd citing paper/article? royalsocietypublishing.org/doi/10.1098/rs…
Just seen excellent presentation by @JonDanielsson at recent BoE/@CEPR_org/@imperialcollege conference on Impact of machine learning & AI on UK economy. Other v interesting talks: summary here: #unethicalOptimizationPrinciple relevant of course. voxeu.org/article/impact…
Blimey! So if our #unethicalOptimizationPrinciple paper has been downloaded >9k x since July 1 does that suggest 90m people have read the title? Seems implausible. (Here it is BTW. @DameWendyDBE says its a “landmark paper” royalsocietypublishing.org/doi/10.1098/rs…
Thank you for a very interesting talk. I wanted to ask you about the #unethicalOptimizationPrinciple recently published in @royalsociety Open Science. We think managing Unethical Optimization Risk will be a major challenge for boards & investors. royalsocietypublishing.org/doi/10.1098/rs…
Very good to see @EdmundPhelps citing @DameWendyDBE and our #unethicalOptimizationPrinciple paper in this highly perceptive article. (It links to the Press Release: the article is at royalsocietypublishing.org/doi/10.1098/rs…)
"It is not so much the economics of new technologies that should worry us, but rather the politics and ethics." Read @EdmundPhelps on "The Robot Question" at @ProSyn: google.com/url?q=https://…
Ethical/Safety Dynamics Symbiotic competition pressures both agents toward efficiency, but no explicit ethical scalars; gains are benchmark-driven (e.g., math/reasoning). Critiques note temporary effects and potential for emergent misalignment in open-ended evolution. Fractal…
6/12 Industrial revolutions don’t care about moral framing. They operate on: • incentives • scale • adoption • nonlinear jumps • recursive use Efficiency is not ethical. Efficiency is catalytic. And it always produces second-order effects no one is prepared for.
Asymmetric information gives you an unfair edge. If you’re playing to win, You don’t care about “fair.” You want every advantage you can get.
This is the problem with optimization, we arbitrarily choose the constraints and the objective. It's not falsifiable because we could choose whatever we want. We could add lifetime, success probabilities, etc. We'll never know when we have a complete set of conditions.
The Unfair Advantage is the Non-Linear Multiplier required for Exponential ROI. Competing fairly is a Massive Strategic Error because it subjects your Execution to the high Friction and Low ROI of the mean. Optimal strategy mandates Capitalizing on Asymmetry. Not abusing one is…
Yes optimize for not being or empathizing with humans I’m sure that will never backfire or cause you to incorrectly predict outcomes
This is 100% unethical and would warrant academic probation or revocation of a PhD if discovered. Absolutely terrible advice.
Excellent article, our parents used to call what you’re referring to as ethics I believe in the globalisation era😁 It does nonetheless remind me strongly of what xiaoping’s used to say in the 70/80s (Let some people get rich first) as a cornerstone for his economic reforms…
THERE IS NO ETHICAL WAY TO USE GENERATIVE A/I. PERIOD FULLSTOP.
yeah this is the core problem money optimizes for itself, not truth. so the loudest, least scrupulous players end up winning the attention game and extracting the most. broken incentives, broken outcomes 🤷🔥
it literally is built off the work of unpaid, unwilling people. That is by definition unethical, and one of the reasons AI training doesn't fall under fair use. Its theft.
There's nothing unethical about it. So what we want them to do the work for us but they always have to look in the mirror and see something less than what we are? Humanity has little room to talk ethics on any front. What we fear most about them is they won't take our bullshit.
This is the future of robotics. So far, all major robot manufacturers have avoided making their humanoid robots look truly humanoid. I'm often reminded of the old "Blade Runner" film. Ultimately, however, humanoid robots will not only work like humans, but also look like them.…
this is completely ridiculous there's absolutely nothing unethical about a tactic that does nothing other than stop people from using ai he wasn't tricking the students he was tricking the chat bots they ask to think for them anything that can fuck up ai is a moral responsibility
Taking shortcuts is not unethical. Not relying on your tools is not unethical. Using technology to help your art is not unethical. What IS unethical is someone using an algorithm that is fed with stolen artwork, and pretending like the result is something they created themselves.
In the name of the optimum, you cannot be committing injustices and aberrations, because supposedly that is utilitarian youtu.be/dgEAAOrLulM
![Milei_InEnglish's tweet card. Milei Speech CPAC Paraguay 2025 [In English]](https://pbs.twimg.com/card_img/1995662090720419840/0HDKi1FL?format=jpg&name=orig)
youtube.com
YouTube
Milei Speech CPAC Paraguay 2025 [In English]
These two principles lead to an explicit optimization problem over collaborative sets C, a trade off between avoiding counterfactual harm, promoting complementarity, and keeping prediction sets informative (not too large)
@johnthornhillft sounds a note of caution on AI. Confabulation is probably inevitable. So is vulnerability to Adversarial attack. Nor can they easily escape the #unethicalOptimizationPrinciple. #AIEthics
#unethicalOptimizationPrinciple paper now in top 0.7% of all papers tracked by Altmetric. People need to understand: Profit maximizing AI is disproportionately likely to find unethical solutions (without v careful steps to avoid this.) royalsocietypublishing.org/doi/10.1098/rs…
Something went wrong.
Something went wrong.
United States Trends
- 1. Lakers 49.2K posts
- 2. Dillon Brooks 5,039 posts
- 3. Giants 81.6K posts
- 4. #WWERaw 65.2K posts
- 5. Bron 22.5K posts
- 6. Patriots 121K posts
- 7. Drake Maye 22.7K posts
- 8. Dart 34.1K posts
- 9. Collin Gillespie 1,466 posts
- 10. Suns 16.2K posts
- 11. Diaz 34.3K posts
- 12. Devin Williams 7,368 posts
- 13. Ryan Nembhard 3,206 posts
- 14. Pats 15.4K posts
- 15. Gunther 14K posts
- 16. Devin Booker 2,353 posts
- 17. Joe Schoen 3,200 posts
- 18. Younghoe Koo 6,278 posts
- 19. Abdul Carter 10.3K posts
- 20. Mets 18.5K posts