brevitea

11 ms·

91.
by brevitea 3y ago
No offense to the author, however both Movie recommendation and Music recommendation algorithms worked quite well before RAG. I actually prefer pre-RAG Music recs. Maybe it's me, but RAG-for-all is giving "we through GenAI at anyt
92.
by brevitea 3y ago
To be sure, there was a former Apple employee and a former Google employee held liable for data exfiltration. But ByteDance is the risk. Okay. Edit: To all the downvoters, check facts vs. hysteria. You know, cuz science and facts and data a
93.
by brevitea 3y ago
Money Laundering.
94.
Who Needs Pegasus When You Have Apple Shortcuts?(jestlandia.github.io)
8 points by brevitea 3y ago|0 comments
95.
by brevitea 3y ago
Perhaps you are seriously underestimating the actual dollars companies spend on advertising.
96.
by brevitea 3y ago
Agree. Similarly, if I pay for a digital copy, I should be able to download that material onto my personal drive.
97.
by brevitea 3y ago
Professor Rafael Núñez (UCSD) wrote a bit about this: https://www.scientificamerican.com/article/how-we-make-sense... .
98.
by brevitea 3y ago
Another favorite: After repeated questions on why Palestinians are starving, CIA Director Bill Burns says Palestinians are starving because "humanitarian assistance can’t get to them." In every response from the Director, the agen
99.
by brevitea 3y ago
Ah, OneTaste, the sex cult. There's a documentary: https://libremdb.iket.me/title/tt22900124 .
100.
by brevitea 3y ago
I did enjoy reading the article using SpeedReader.
101.
by brevitea 3y ago
Sure wish CISA and SEC would effectively monitor and fine companies that suffer data breaches. After all, we're not being paid for that data, yet we remain the victim of their actions.
102.
by brevitea 3y ago
IMO, the more the merrier. That's the joy of decentralization and P2P.
103.
by brevitea 3y ago
Will the CEO be held liable? Will their yearly multi-million dollar bonuses be taken?
104.
by brevitea 3y ago
Thanks for the demo video; very cool. How do you prevent prompt injection attacks? E.g., if you create a user account for the LLM models to run in, how to do prevent an attack where the LLMs can be leveraged to execute privilege escalation