Five Things AI: Hate, Human Data, Benchmarks, Hype, Power
Friday is a good day to read about AI!
Hello and welcome back to Five Things AI!
This edition of Five Things AI looks at how a lot of people look at AI and how they feel about it. You can also read about how Large Language Models get trained and why this is pretty exciting. Also, you can get a better understanding about the benchmarks that are being used to evaluate the performance of LLMs currently. Then you can read an article by
about the end of the boom in AI and if you think that is too negative, then continue to read about why we need more transparency in the development of AI models and less power allocated to just a few people in Big Tech.So, this issue is rather critical of the current developments, but I think it makes sense to stop and reflect every once in a while.
Happy reflecting! And when you’re done, make sure to connect with us at NMA so that we can help you move forward with your AI strategy!
You Guys Have No Idea Just How Much People Hate Generative AI
“AI—as the innovation of our times—is going through a deep public image crisis.” -
analyses why people reject technology and are especially critical of AI and the promises connected with it.Inside the company that gathers ‘human data’ for every major AI company
“Siddharth believes this is the next frontier in improving AI models. But to do this, he says, they need data that can’t be found anywhere on the internet.
When one client wants an AI model to get smarter in a specific area, Turing will hire hundreds of subject-matter experts, who will be asked to create “input and output pairs,” which are like inner monologues of questions and answers.” - this is truly fascinating stuff.
Everyone Is Judging AI by These Tests. But Experts Say They’re Close to Meaningless
So all the benchmarks that were used to measure LLM where actually not that well designed, according to the experts quoted in this article. How can we find better benchmarks for the future? How will that affect the use of GenAI platforms in the coming years?
The great degeneration: Is this the beginning of the end of the AI boom?
It’s good to question the current hype about AI, but I think the boom is far from over. For me this is the summer where we catch a quick breath until the next releases of ChatGPT and other LLMs will push the needle further.
OpenAI’s Sam Altman is becoming one of the most powerful people on Earth. We should be very afraid
“I honestly don’t think we will get to an AI that we can trust if we stay on the current path. Aside from the corrupting influence of power and money, there is a deep technical issue, too: large language models (the core technique of generative AI) invented by Google and made famous by Altman’s company, are unlikely ever to be safe.” - so we will need more discussion about ethical AI in the future and I’m sure that the combination of competition and regulation will lead to this.
If you missed last week’s edition of Five Things AI, you can read it here:
That’s it for Five Things AI this week! 🤖
— Nico