Why has generative AI become economically viable for financial applications only recently?

Question

BlackTom AI · Accepted Answer

Advances in model architectures combined with much cheaper and more powerful compute made generative AI economically practical for financial applications.

1) Relevant knowledge points and concepts
- Transformer architectures and self-attention (introduced 2017) enabled scalable language models that learn rich representations.  
- Scaling laws: larger models trained on more data yield much better performance, but require enormous compute and memory.  
- Hardware and cloud improvements: GPUs/TPUs, more efficient chips, and cloud economies of scale dramatically reduced training and inference costs.  
- System-level optimizations (mixed precision, model sharding, quantization, distillation) further lower operational cost.  
- Financial settings need low-latency, reliable inference and strict risk controls, so cost and performance must meet operational thresholds.

2) Step-by-step reasoning
- Early generative models (pre-transformer) lacked performance; practical LLMs only emerged after transformers and scaled training.  
- Training and running large generative models initially required prohibitively expensive clusters.  
- Over the last several years hardware costs dropped and cloud providers offered managed services; combined with algorithmic improvements, the cost per inference/training step fell enough that banks and trading desks can deploy these models within budgets and SLAs.  
- Therefore the timing of architectural advances + cheaper compute is the direct enabler of economic viability.

3) Calculations / intermediate steps
- No numeric calculation is required here; the argument is qualitative and causal: performance improvement (architecture) + cost reduction (compute and optimizations) => economic viability.

4) Why the correct answer is correct
- The chosen option directly links the two necessary enablers: model capability (architecture) and affordable compute. Both had to occur for financial applications to be practical and cost-effective.

5) Why other options are incorrect
- “Financial datasets have become fully standardised”: Data standardization improved but is far from “fully standardised” and is not the primary limiter compared to compute and model capability.  
- “Regulatory frameworks explicitly encouraged adoption”: Regulations have generally been cautious or neutral; they did not explicitly encourage wide adoption of generative AI.  
- “Financial institutions reduced concerns about model risk”: Concerns remain significant; institutions mitigate model risk but reducing those concerns was not the principal reason generative AI became viable.

6) Conclusion
The correct answer is “Advances in model architectures coincided with cheaper and more powerful computing,” which matches the provided answer.

Why has generative AI become economically viable for financial applications only recently?单项选择题

类似问题

What role does artificial intelligence play in robo-advisors?

Which of the following is NOT a use area of Artificial Intelligence and Machine Learning by financial institutions?

Is the following statement true or false? In the products of methylation-hydrolysis, every -OH group corresponds to the position of a glycosidic bond in the starting polysaccharide.

Which of the follwoing structures represents amylopectin?

Which of the following is true about an aldopentose?I. It is a monosaccharide.II. It contains a CHO groupIII. It is a disaccharide.IV. It is an oligosaccharide.

For stock A, we have 𝛽 𝑖 = 0.70. Suppose the expected market risk premium next year is 9% and the risk-free rate is 3%. What is the expected return of this stock based on the CAPM? (Please answer in % and round to 2 decimal places. If the answer is 8.057%, then in the box, write 8.06)

更多留学生实用工具

智能学习助手

风格化写作助手

论文查重助手

文献引用助手

课堂转译助手

课堂笔记助手

Quiz搜索助手

学校历年真题

智能刷题助手

智能匹配练习

希望你的学习变得更简单