Synthetic intelligence (AI) won’t ever grow to be a acutely aware being on account of a scarcity of intention, which is endemic to human beings and different organic creatures, based on Sandeep Nailwal — co-founder of Polygon and the open-source AI firm Sentient.
“I do not see that AI could have any important degree of conscience,” Nailwal advised Cointelegraph in an interview, including that he doesn’t imagine the doomsday state of affairs of AI turning into self-aware and taking on humanity is feasible.
The manager was crucial of the idea that consciousness emerges randomly on account of advanced chemical interactions or processes and mentioned that whereas these processes can create advanced cells, they can not create consciousness.
As an alternative, Nailwal is worried that centralized establishments will misuse synthetic intelligence for surveillance and curtail particular person freedoms, which is why AI have to be clear and democratized. Nailwal mentioned:
“That’s my core thought for a way I got here up with the thought of Sentient, that finally the worldwide AI, which might truly create a borderless world, must be an AI that’s managed by each human being.”
The manager added that these centralized threats are why each particular person wants a customized AI that works on their behalf and is loyal to that particular particular person to guard themselves from different AIs deployed by highly effective establishments.
Sentient’s open mannequin strategy to AI vs the opaque strategy of centralized platforms. Supply: Sentient Whitepaper
Associated: OpenAI’s GPT-4.5 ‘gained’t crush benchmarks’ however is likely to be a greater buddy
Decentralized AI can assist stop a catastrophe earlier than it transpires
In October 2024, AI firm Anthropic launched a paper outlining eventualities the place AI might sabotage humanity and potential options to the issue.
Finally, the paper concluded that AI just isn’t a right away risk to humanity however might grow to be harmful sooner or later as AI fashions grow to be extra superior.
Various kinds of potential AI sabotage eventualities outlined within the Anthropic paper. Supply: Anthropic
David Holtzman, a former army intelligence skilled and chief technique officer of the Naoris decentralized safety protocol, advised Cointelegraph that AI poses an enormous danger to privateness within the close to time period.
Like Nailwal, Holtzman argued that centralized establishments, together with the state, might wield AI for surveillance and that decentralization is a bulwark towards AI threats.
Journal: ChatGPT set off pleased with nukes, SEGA’s 80s AI, TAO up 90%: AI Eye