#unethicaloptimizationprinciple 搜尋結果
@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…
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…
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…
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://…
@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
Not least because of the #unethicalOptimizationprinciple It’s now an established scientific fact that profit-maximising AI disproportionately tends to find unethical strategies. royalsocietypublishing.org/doi/10.1098/rs…
This should be v interesting. Dr Lee have you seen the #unethicalOptimizationPrinciple which @DameWendyDBE says is a “landmark paper”? I’d be v interested in your thoughts. royalsocietypublishing.org/doi/10.1098/rs…
V interesting book judging from review in Spectator by @doctorow. An additional aspect is the recently discovered #unethicalOptimizationPrinciple that AI to optimize returns will disproportionately pick unethical strategies ☹ royalsocietypublishing.org/doi/10.1098/rs…
What makes it more dangerous is the #unethicalOptimizationPrinciple royalsocietypublishing.org/doi/10.1098/rs…
Worth bearing in mind the #unethicalOptimizationPrinciple which tends to make things worse. royalsocietypublishing.org/doi/10.1098/rs…
Bear in mind recently discovered #unethicalOptimizationPrinciple: If an AI aims to maximize risk-adjusted return, it's disproportionately likely (under mild conditions) to pick an unethical strategy unless objective function allows sufficiently for this. doi.org/10.1098/rsos.2…
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://…
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…
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…
Extended #unethicalOptimizationPrinciple in action 🙁 royalsocietypublishing.org/doi/10.1098/rs…
Very interesting. But we need to bear in mind, & guard against, the recently proven #unethicalOptimizationPrinciple: AI that aims to maximise profits/returns is disproportionately likely to pick unethical strategies. 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…
BTW I think an extended version of the #unethicalOptimizationPrinciple, that doesn't restrict itself to the provable case of AI, underlies many of the problems you mention. royalsocietypublishing.org/doi/10.1098/rs…
Indeed, & UK can take a lead in this. We need to bear in mind #unethicalOptimizationPrinciple which proves(!) that AI systems have to have responsibility designed in carefully otherwise they disproportionately pick unethical strategies. royalsocietypublishing.org/doi/10.1098/rs…
#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…
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…
You might find the #unethicalOptimizationPrinciple relevant to your argument. royalsocietypublishing.org/doi/10.1098/rs…
Need to bear in mind #unethicalOptimizationPrinciple royalsocietypublishing.org/doi/10.1098/rs…
And, dare I say it, we need people to be aware of #unethicalOptimizationPrinciple so they are on the alert for the (now mathematically established fact) that AI which tries to profit maximise will tend to disproportionately pick unethical solutions.
@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. Cyber Monday 49.4K posts
- 2. GreetEat Corp. N/A
- 3. TOP CALL 11.6K posts
- 4. Adam Thielen 2,665 posts
- 5. Shakur 6,244 posts
- 6. #Rashmer 19K posts
- 7. Admiral Bradley 1,542 posts
- 8. Alina Habba 34.5K posts
- 9. #GivingTuesday 3,012 posts
- 10. #IDontWantToOverreactBUT 1,501 posts
- 11. Hartline 2,945 posts
- 12. Check Analyze N/A
- 13. Token Signal 3,756 posts
- 14. $MSTR 18.1K posts
- 15. Shopify 4,962 posts
- 16. Marty Supreme 3,405 posts
- 17. UCLA 7,177 posts
- 18. Market Focus 3,015 posts
- 19. #MondayMotivation 10.8K posts
- 20. John Butler N/A