Life was going smoothly, but suddenly, everything seemed related to Meta, AI, Instagram, Twitter, or TikTok. Meta has formed the Open Innovation AI Research Community to jump on the AI bandwagon and demonstrate that they are still relevant in this tech-savvy era. As they describe, the big objective is fostering openness, creativity, and cooperation among AI researchers. Doesn't it sound impressive? Hold your breath because when we go further, the truth may be less "wow" and more "meh."
Initially, this community will focus on the privacy, safety, and security of well-known language models, such as OpenAI's ChatGPT. They also want to share their thoughts on improving AI models and developing a research agenda for the future. They say their own researchers will be included in the gang, but this so-called "member-led" community will be in charge. Meta AI, Meta's AI R&D department, will just be a glorified facilitator - talk about hands-off parenting!
Meta pledges to host seminars on "critical open research questions" and to provide rules for appropriate open-source model creation and distribution. That's all well and good, but where are the specifics? They may someday have a website, social networks, and academic research submissions, but it's just "we'll see."
But hey, wait! There is always more to the story. You are missing the best part; this is when things start to get interesting. This research community's members are expected to support their own work. Yes, you read that correctly. Meta, the tech titan, will not spare this community a dime or byte of processing power. They presumably don't want to be accused of undue influence, but that's like holding a dinner party and making your guests pay for the food!
Let's go back in time for a bit. Meta's AI adventures haven't always gone well. Remember when they had to cancel an AI demonstration because it turned out to be a tiny racist? Oops. Their AI ethics committee? Toothless. They released anti-AI bias tools? Completely ineffective. And don't even get us started on how their algorithms have been accused of increasing socioeconomic inequality and bigotry toward Black people.
They are now attempting to win the hearts of "professors at accredited universities" with "relevant experience with AI." But here's the thing: many open machine-learning research communities need to march to the beat of Big Tech. So, what makes these academics want to participate in the Meta show?
Sure, we may be mistaken. Perhaps, just perhaps, Meta's Open Innovation AI Research Community will surprise us all by delivering on its promise of robust and representative models. However, we are suspicious since the paucity of resources committed from the beginning does not inspire much trust.
You have until September 10 to apply if you want to join this AI adventure. They promise to embrace varied research fields and technical skills, but the judgment is yet out on whether this will be a huge platform for AI marvels or simply another show to put on the "Meta."
Read next: This New Report Reveals the World’s Divided Opinions on AI
Initially, this community will focus on the privacy, safety, and security of well-known language models, such as OpenAI's ChatGPT. They also want to share their thoughts on improving AI models and developing a research agenda for the future. They say their own researchers will be included in the gang, but this so-called "member-led" community will be in charge. Meta AI, Meta's AI R&D department, will just be a glorified facilitator - talk about hands-off parenting!
Meta pledges to host seminars on "critical open research questions" and to provide rules for appropriate open-source model creation and distribution. That's all well and good, but where are the specifics? They may someday have a website, social networks, and academic research submissions, but it's just "we'll see."
But hey, wait! There is always more to the story. You are missing the best part; this is when things start to get interesting. This research community's members are expected to support their own work. Yes, you read that correctly. Meta, the tech titan, will not spare this community a dime or byte of processing power. They presumably don't want to be accused of undue influence, but that's like holding a dinner party and making your guests pay for the food!
Let's go back in time for a bit. Meta's AI adventures haven't always gone well. Remember when they had to cancel an AI demonstration because it turned out to be a tiny racist? Oops. Their AI ethics committee? Toothless. They released anti-AI bias tools? Completely ineffective. And don't even get us started on how their algorithms have been accused of increasing socioeconomic inequality and bigotry toward Black people.
They are now attempting to win the hearts of "professors at accredited universities" with "relevant experience with AI." But here's the thing: many open machine-learning research communities need to march to the beat of Big Tech. So, what makes these academics want to participate in the Meta show?
Sure, we may be mistaken. Perhaps, just perhaps, Meta's Open Innovation AI Research Community will surprise us all by delivering on its promise of robust and representative models. However, we are suspicious since the paucity of resources committed from the beginning does not inspire much trust.
You have until September 10 to apply if you want to join this AI adventure. They promise to embrace varied research fields and technical skills, but the judgment is yet out on whether this will be a huge platform for AI marvels or simply another show to put on the "Meta."
Read next: This New Report Reveals the World’s Divided Opinions on AI
