yennie jun
@artfish.bsky.social
ai research @GoogleDeepMind || prev. @Microsoft, @UniofOxford, @UNGlobalPulse, @DeepLearningAI_
my blog: artfish.ai
my blog: artfish.ai
whether it's using ai to automate the scientific process or to take the role of judges/therapists/doctors/artists, it's important to remember that ai is here to serve humans — not the other way around!
December 31, 2024 at 10:48 AM
whether it's using ai to automate the scientific process or to take the role of judges/therapists/doctors/artists, it's important to remember that ai is here to serve humans — not the other way around!
at the end of the day, whether it's 2024 or 2025, it is important not to lose sight of the **human** within the system!! 💙
December 31, 2024 at 10:48 AM
at the end of the day, whether it's 2024 or 2025, it is important not to lose sight of the **human** within the system!! 💙
some of the things i am most excited for 2025!
- development of more difficult benchmarks and evaluations for AI systems
- ongoing questions and discussions about alignment and safety
- and yeah, agents (despite the over-hype, i do think there is a lot of value and substance)
- development of more difficult benchmarks and evaluations for AI systems
- ongoing questions and discussions about alignment and safety
- and yeah, agents (despite the over-hype, i do think there is a lot of value and substance)
December 31, 2024 at 10:48 AM
some of the things i am most excited for 2025!
- development of more difficult benchmarks and evaluations for AI systems
- ongoing questions and discussions about alignment and safety
- and yeah, agents (despite the over-hype, i do think there is a lot of value and substance)
- development of more difficult benchmarks and evaluations for AI systems
- ongoing questions and discussions about alignment and safety
- and yeah, agents (despite the over-hype, i do think there is a lot of value and substance)
some of the patterns i observed within 2024 that i found impressive and compelling:
- a flourishing of smaller, cheaper, and more efficient models 🐥
- reasoning models that use more inference compute (thinking and agentic models)
- ai-generated video (really blows my mind)
- a flourishing of smaller, cheaper, and more efficient models 🐥
- reasoning models that use more inference compute (thinking and agentic models)
- ai-generated video (really blows my mind)
December 31, 2024 at 10:48 AM
some of the patterns i observed within 2024 that i found impressive and compelling:
- a flourishing of smaller, cheaper, and more efficient models 🐥
- reasoning models that use more inference compute (thinking and agentic models)
- ai-generated video (really blows my mind)
- a flourishing of smaller, cheaper, and more efficient models 🐥
- reasoning models that use more inference compute (thinking and agentic models)
- ai-generated video (really blows my mind)
i first discuss the contradiction that i find within ai research:
- that there is so much new and amazing research happening all of the time; - and yet so much of this research hasn't reached the public;
- and at the same time, a portion of the research is already outdated 😱
- that there is so much new and amazing research happening all of the time; - and yet so much of this research hasn't reached the public;
- and at the same time, a portion of the research is already outdated 😱
December 31, 2024 at 10:48 AM
i first discuss the contradiction that i find within ai research:
- that there is so much new and amazing research happening all of the time; - and yet so much of this research hasn't reached the public;
- and at the same time, a portion of the research is already outdated 😱
- that there is so much new and amazing research happening all of the time; - and yet so much of this research hasn't reached the public;
- and at the same time, a portion of the research is already outdated 😱