Roman is so smart, down to earth, humble. He talks so clearly, without any mambo jumbo and buzzwords. A truly passionate and dedicated professional. (Even thought I am biased, as he is from my hometown, Riga, Latvia )
Great dialogue 👌 I agree it is the definition of trials & tribulations for many past generations. But what do you do when everyone refused cut n dry programs cried for ai nepotism instead of properly innovating along the way ? Many of them made the avalanches of tech gatekeeping perks or position of authority . I read Darpa say they knew more than the current capabilities. The next chip update gets tougher, but it is 80 yrs into the transitor age. Microchips have been on island far from American domestic courts jurisdiction for a long time. The attempts to censor and block information like anthropic assuming everyone is a car theif instead of individual liberal entrepreneurship who wants startup biz unlocking cars or just save money as multi skillset is the cause of nilhisms and hulucinations I've come across from a far. Timed on well established no hold bar proven phylosphical lines of measure like elon appears to be doing with longitude latitude pragmatic common sense objectivism proper gen x platform, self driver, optimus hooked up to Austin. Xai has to be completely differently tuned for science
Thanks both! If your audience would like to hear more from Roman I was lucky enough to interview him for @Sentientism about "what's real?", "what matters?" "who matters?" and "how can we make a better future?" Hopefully a good complement to your conversation: ua-cam.com/video/SK9lGvGmITc/v-deo.html
I don't visualize SAI as a "supermind" with independent thought & intentions. However, a huge problem with SAI may be errors of translation/interpretation in communications - particularly with its human counterparts. I foresee this because I presume SAI's evolution will "naturally" involve creating for itself a super level high tech language (SLHTL) to speed up all its communications, calculations & programming in conjunction with other hosts or devices (either design-engineered by humans or by SAI itself). SLHTL speed would evolve to minimize both energy consumption & heat build-up in its operating systems. Once created the interface of SLHTL with human language would likely result in translation errors in either direction which could have dangerous or fatal consequences if applied to safety-critical procedures & operations. The speed at which any semi-autonomous SAI would operate means it would be impossible for humans to monitor SAI decisions in real time & so any SAI "hands free" operations would inherently be very risky.
When can we expect AI/SAI will be able to innovate & engineer chemistry solutions to break down or modify the covalent bond of CO2 so that it can, say, be bound to aluminium ore and sequestered or repurposed ?
This is fire 🔥🔥🔥🔥… Enlightening and Practical knowledge. Thank you for sharing this wisdom. So grateful for the Foresight Institute and Roman Yampolskiy for this epoch progress knowledge. Fantastic questions and answers!!
Roman is so smart, down to earth, humble. He talks so clearly, without any mambo jumbo and buzzwords. A truly passionate and dedicated professional. (Even thought I am biased, as he is from my hometown, Riga, Latvia )
Great dialogue 👌
I agree it is the definition of trials & tribulations for many past generations. But what do you do when everyone refused cut n dry programs cried for ai nepotism instead of properly innovating along the way ? Many of them made the avalanches of tech gatekeeping perks or position of authority .
I read Darpa say they knew more than the current capabilities. The next chip update gets tougher, but it is 80 yrs into the transitor age. Microchips have been on island far from American domestic courts jurisdiction for a long time.
The attempts to censor and block information like anthropic assuming everyone is a car theif instead of individual liberal entrepreneurship who wants startup biz unlocking cars or just save money as multi skillset is the cause of nilhisms and hulucinations I've come across from a far.
Timed on well established no hold bar proven phylosphical lines of measure like elon appears to be doing with longitude latitude pragmatic common sense objectivism proper gen x platform, self driver, optimus hooked up to Austin.
Xai has to be completely differently tuned for science
Sorry but can't help thinking of Rasputin when seeing Roman ! On the plus side, Roman is a force for good, which Rasputin was not.
Gratis. His words a fresh breeze after the tiresome huffed hopium often found on Foresight.
Thanks both! If your audience would like to hear more from Roman I was lucky enough to interview him for @Sentientism about "what's real?", "what matters?" "who matters?" and "how can we make a better future?" Hopefully a good complement to your conversation: ua-cam.com/video/SK9lGvGmITc/v-deo.html
I don't visualize SAI as a "supermind" with independent thought & intentions. However, a huge problem with SAI may be errors of translation/interpretation in communications - particularly with its human counterparts. I foresee this because I presume SAI's evolution will "naturally" involve creating for itself a super level high tech language (SLHTL) to speed up all its communications, calculations & programming in conjunction with other hosts or devices (either design-engineered by humans or by SAI itself).
SLHTL speed would evolve to minimize both energy consumption & heat build-up in its operating systems. Once created the interface of SLHTL with human language would likely result in translation errors in either direction which could have dangerous or fatal consequences if applied to safety-critical procedures & operations.
The speed at which any semi-autonomous SAI would operate means it would be impossible for humans to monitor SAI decisions in real time & so any SAI "hands free" operations would inherently be very risky.
When can we expect AI/SAI will be able to innovate & engineer chemistry solutions to break down or modify the covalent bond of CO2 so that it can, say, be bound to aluminium ore and sequestered or repurposed ?
The intro host AI generated ?
Man, Roman does not sugar coat his position. The only chance we have is IF he is wrong
This is fire 🔥🔥🔥🔥… Enlightening and Practical knowledge. Thank you for sharing this wisdom. So grateful for the Foresight Institute and Roman Yampolskiy for this epoch progress knowledge. Fantastic questions and answers!!