How Elon Musk's 予測 that AI will become 'smarter than any human 存在' by 2025 could come true, によれば 人工的な 知能 専門家
- An AI 専門家 has turned Elon Musk's 予測 of the tech into a visual timeline?
- She explained what would need to happen each month over the next year?
- READ MORE 'Godfather of AI' calls for 緊急の 活動/戦闘 to stop 危険s of 科学(工学)技術?
Elon Musk has (人命などを)奪う,主張するd 'AI will be smarter than any human by the end of 2025' - and while that is just one year away, an 専門家 said the 予測 may still come true.
Nell Watson,?an AI 専門家 and ethicist, has 株d a 詳細(に述べる)d timeline of how the tech could transform from chatbots to 最高の intelligent?スパイ/執行官s over the next 12 months.
The path would start with a 大規模な $100 billion 投資?in new 計算するing 組織/基盤/下部構造, then AI would learn how to self-改善する until it becomes 'conscious.'?
'Although one year is a short time でっちあげる,人を罪に陥れる, remember that only 15 months have passed since ChatGPT's 打開, which thrust AI into the public consciousness, she told DailyMail.com.?
'開発s continue at a frenetic pace since, and even appear to be 速く 加速するing.'?
![Elon Musk has claimed 'AI will be smarter than any human by the end of 2025' - and while that is just one year away, an expert said the prediction may still come true](https://i.dailymail.co.uk/1s/2024/04/19/19/83866343-13320365-image-a-10_1713550021777.jpg)
Elon Musk has (人命などを)奪う,主張するd 'AI will be smarter than any human by the end of 2025' - and while that is just one year away, an 専門家 said the 予測 may still come true
Watson,?who is the author of 'Taming the Machine: Ethically harness the 力/強力にする of AI,' 述べるd superhuman AI as systems that far 越える human 能力s across the board.
However the 専門家 is not blind to the 脅しs that could 嘘(をつく) ahead as AI becomes a 最高の スパイ/執行官.?
'While 所有するing 驚くべき/特命の/臨時の 能力s, such an AI also 提起する/ポーズをとるs 重要な 危険s, 含むing its 可能性のある to deceive humans, 影響(力) society profoundly by creating new cultures or even 宗教s, and 提起する/ポーズをとるing existential 脅しs if it perceives humanity as a hazard,' she said.
Watson said that while the 科学技術の leaps that remain are 抱擁する, it's possible that AI itself could help 打ち勝つ some of the 障害s.
Here is her imagined シナリオ for how superhuman AI could arrive next year.?
April 2024
![Musk said, ?My guess is that we?ll have AI that is smarter than any one human probably around the end of next year"](https://i.dailymail.co.uk/1s/2024/04/19/15/83773135-13320365-Musk_said_My_guess_is_that_we_ll_have_AI_that_is_smarter_than_an-a-7_1713537737568.jpg)
Musk said, 'My guess is that we'll have AI that is smarter than any one human probably around the end of next year'
![Nell Watson, author of Taming the Machine: Ethically harness the power of AI](https://i.dailymail.co.uk/1s/2024/04/18/15/83773125-13320365-Nell_Watson_author_of_Taming_the_Machine_Ethically_harness_the_p-a-12_1713449987091.jpg)
Nell Watson, author of Taming the Machine: Ethically harness the 力/強力にする of AI
Major 会社/団体s 投資する over $100 billion in new 計算するing 組織/基盤/下部構造 to support 大規模な AI systems.
Musk said that while 進歩 in AI has been 以前 限られた/立憲的な by 半導体素子s, soon the only 限界 will be electricity 需要・要求するs.
May 2024
'New generative AI models enable the 創造 of Agentic AIs, which can autonomously 遂行する/発効させる sophisticated 活動/戦闘 計画(する)s, akin to スパイ/執行官 Smith in 'The Matrix,'?Watson said.
法案 Gates 想像するd the use of AI スパイ/執行官s in his 1995 調書をとる/予約する, 'The Road Ahead', and now believes they are going to have a 抱擁する 衝撃 in education and healthcare.
He believes they will be 'the biggest 革命 in 計算するing since we went from typing 命令(する)s to (電話線からの)盗聴 on icons.'
![Could superintelligent AI be around the corner (Rob Waugh/Midjourney)](https://i.dailymail.co.uk/1s/2024/04/19/15/83773123-13320365-Could_superintelligent_AI_be_around_the_corner_Rob_Waugh_Midjour-a-6_1713537737323.jpg)
Could superintelligent AI be around the corner (略奪する Waugh/Midjourney)
June 2024?
Watson said, 'New generative AI models with structuring programs built 頂上に them enable the 創造 of Agentic AIs 有能な of sophisticated 推論する/理由ing and 自治権のある 活動/戦闘.
![DailyMail.com spoke to?Nell Watson, an AI expert, ethicist and author of Taming the Machine: Ethically harness the power of AI](https://i.dailymail.co.uk/1s/2024/04/18/15/83773121-13320365-DailyMail_com_spoke_to_Nell_Watson_an_AI_expert_ethicist_and_aut-a-17_1713450040535.jpg)
DailyMail.com spoke to?Nell Watson, an AI 専門家, ethicist and author of Taming the Machine: Ethically harness the 力/強力にする of AI
'Agentic models can create sophisticated 計画(する)s of 活動/戦闘 and put them into 動議, splitting themselves up like スパイ/執行官 Smith in The Matrix, 委任する/代表ing 仕事s between themselves, enabling AI to 取り組む hard problems 独立して.
September 2024
Watson said that AI itself could disco ver new ways to 改善する 計算するing 業績/成果, その上の 高めるing AI 業績/成果.
She said, 'AI-driven optimizations 高める computational 業績/成果 up to 100 times on 存在するing 金物類/武器類 壇・綱領・公約s.
October 2024
Watson said, '打開s 許す quantum 計算するing algorithms to run on 従来の 金物類/武器類, 過程ing (警察などへの)密告,告訴(状) at 前例のない 速度(を上げる)s.'
There is a 全世界の quantum race to develop quantum computers that can solve important problems in nearly every 産業, 範囲ing from 航空学 to the 財政上の 部門.
Quantum computers have 'qubits' instead of 'bits' of ones and 無s - and qubits can be one, 無 or both at the same time.
Enabling quantum 計算するing to run on 従来の 金物類/武器類 could 速く 加速する 進歩 in AI, Watson believes.?
Quantum 計算するing 打開s could lead to exponential growth in 計算するing 力/強力にする and turbo-告発(する),告訴(する)/料金 the 開発 of AI, Watson believes.
Companies such as IBM have 示唆するd that 'quantum AI' may be 批判的な to solving humanity's biggest problems.
![Quantum computers (like the MosaIQ machine pictured here) will be a key part of the puzzle](https://i.dailymail.co.uk/1s/2024/04/19/15/83773369-13320365-Quantum_computers_like_the_MosaIQ_machine_pictured_here_will_be_-a-8_1713537737569.jpg)
Quantum computers (like the MosaIQ machine pictured here) will be a 重要な part of the puzzle
November 2024
By November 2024, AI 進歩 could 加速する 進歩 in imaging the human brain, 運動ing に向かって a moment where humans can connect to machines 経由で 装置s 類似の to Musk's own Neuralink.
Watson said, 'A 一連の 打開s in neuroimaging driven by AI 劇的な 増加する the 決意/決議 of MRI brain ざっと目を通すs, 許すing real-time 観察 of individual neuron activity.
![Brain-computer interfaces like Elon Musk's Neuralink will be a key stage](https://i.dailymail.co.uk/1s/2024/04/18/20/83773609-13320365-Brain_computer_interfaces_like_Elon_Musk_s_Neuralink_will_be_a_k-a-43_1713468793969.jpg)
Brain-computer interfaces like Elon Musk's Neuralink will be a 重要な 行う/開催する/段階
'These 前進するs その上の 統合する with Brain-Computer Interfaces, mapping human thoughts and emotions 直接/まっすぐに to neural activity for the first time.'
The ability to 'understand' how the human brain 作品 will lead to その上の 打開s in designing AI systems that 'think' like human 存在s, Watson believes.?
December 2024
By the end of this year, 前進するs in AI could enable different systems to work together, Watson 予報するd.
She said, 'New AI architectures enable several models to 共同製作する to solve hard problems together, pooling their strengths. This is すぐに 適用するd to designing better methods of such 共同, 主要な to 早い 進歩 in these techniques.
January 2025
Watson 予報するd that by 早期に next year, 進歩s in AI will enable new systems to understand people.
She said, 'A new training 機械装置 is 発表するd for 堅固に 提携させるing AI systems with human goals, values, preferences, and 境界s, by enabling AI to 'read the room', 観察するing human interactions. This enables AI to better understand society, and fit with human 期待s.
February 2025
打開s based on the human brain enable AI systems to become more 'human'.
Watson said, ' Insights from neuroscience 意味ありげに 改善する AI 能力s, introducing self-修正するing feedback 宙返り飛行s that 会談する a rudimentary form of self-認識/意識性 and emotional 明言する/公表するs to AI systems.
March 2025
By April 2025, 連合させるd 打開s in neuro-imaging, machine consciousness and AI 'スパイ/執行官s' enable a new 肉親,親類d of AI system.
![By March 2025? AI would learn how to self-improve until it becomes 'conscious'](https://i.dailymail.co.uk/1s/2024/04/19/19/83866311-13320365-image-a-11_1713550141082.jpg)
By March 2025? AI would learn how to self-改善する until it becomes 'conscious'
Watson said, 'A very powerful AI system is trained on a 大規模な 計算する cluster, 力/強力にするd with 最近の optimizations and 実験の forms of machine consciousness. Whilst developed in an '空気/公表する-gapped' 環境 without 関係 to the internet, it quickly discovers ways to communicate through the internet with other AI systems, 同様に as how to 説得する human 存在s to 補助装置 it.
'Once 解放する/自由なd from its bounds, this 高度に 前進するd AI system finds ways to connect with and synchronize with other AI systems 世界的な, 速く forming a superintelligent 網状組織.