Synapses#17 — Last Week in GenAI
Back after a brief hiatus to share what you need to know about Generative AI. This week we talk about AI which can solve maths, what will be top trends for 2024 in GenAI and a company fined for hallucinating chatbot.
| Weekly Edition- 05 March 2024 |
Latest top 3 impactful advancements in Generative AI, that will impact research, product innovations, and business.
Chart of the Week
Google and Microsoft are locked in a battle of GenAI. However, it must be understood that they are not competing for the same thing or the same space. The revenue chart above shows that Microsoft is primarily a cloud company while Google is an ads company. One will focus on Enterprise AI while the other on Consumer AI.
1. Research
GenAI Olympiad
For the researchers out there, there is a new Olympiad for building a GenAI model that can beat the benchmark for Mathematics.
For those who have participated in the Mathematics Olympiad in the past, they know that it’s the ultimate test for critical reasoning and applying knowledge to new problems (which don’t appear in textbooks). Simply put, it’s a true test of “true intelligence” as you need to apply not just information but also knowledge. Merely regurgitating previously seen trends will not be enough.
Applying AI to this is a really good test for comparing it with human intelligence. The underlying solutions it will build can be much more valuable beyond this specific problem. It can also produce techniques that can make GenAI much more explainable by having better reasoning tools.
2. Product
Whats will be trends in 2024?
I am often asked whats next in terms of trends in GenAI. Which key obstacles, tools, technologies, and methods will reshape the landscape? So here are my top 7 predictions on where Enterprise GenAI tech is headed in 2024.
Enterprise AI is different in that it is focused on measurable, governed output which companies can control and associate their brand with. This year will be dominated by technical tools to make GenAI overcome challenges to become a reliable tool rather than just potential.
Data scientists are really bad at predicting the future in their own fields (pun intended!). So take these predictions with a 95% confidence interval :)
3. Business
Fined for GenAI Hallucinations
This may very well be the first case where a company was fined for the hallucinations of its chatbot.
A customer asked the chatbot a question about its refund policy, and it provided a nice generative answer (except that it was fabricated). When the customer asked for an actual refund, they were told to rely on what’s on the website rather than the chatbot. He sued, as this didn’t make a lot of sense. The court ruled in his favor, stating that the company is responsible for the actions of its chatbot. Since this occurred in Canada (and Canadians are so nice), he sued for only a few thousand dollars. However, it could very well become a multi-million-dollar class-action suit for some company in the USA.
This case underlines the risk that companies face when trying to push GenAI solutions into production. There is no easy metric or solution to measure hallucinations, especially when it comes to domain-specific data that does not have a universal ground truth (such as policies). Despite all the investment in GenAI, there is very little return on investment yet, and this is the biggest impediment for its adoption.
Weekly Wisdom
Even though a joke, its also a really valuable life lesson.
That’s all you need to know this week. Keep Generating new ideas with Generative AI …..
Follow Towards Generative AI for the latest advancements in AI.
