Thoughts on all this AI stuff?
Rumors of GPT-4's release are on Twitter rn, and all these AI companies are raising crazy funds like Jasper or Stability AI. How do you see these companies affecting finance/life more generally? It seems like with publicly available models, lots of new technologies are popping up left & right
I don't really have thoughts of my own here, but there are many really smart people with interesting takes on AI. OP you are likely familiar with these people if you know about GPT-4, but for the rest of WSO: Ray Kurzweil & Ben Goertzel (the optimists), Sam Harris & Nick Bostrom (both really scared), Sam Altman & Demis Hasabis (the innovators), hell even Elon Musk. To put it crudely, the consensus among these types is that shit's gonna get real fuckin weird, real fuckin fast. TBD if weird in a good way or a bad way.
Dead internet theory.
I think AI is way overhyped. It's one of those things that Silicon Valley loves to window dress and market as being a game-changer that will "disrupt" the world when really 90% of Machine Learning/AI that I've seen is essentially just glorified statistics. I do think that there will be a lot more tools that can automate some of the grunt work in the white collar world, which will impact jobs, but I also think that work will evolve and technological progress will just mean the work will become more advanced/analytical. It's kind of like how the spreadsheet didn't get rid of finance/accounting professionals or even reduce the workload. The expectation has now become that we can do much more advanced analysis that simply wasn't possible back in the day when just calculating and tying out numbers took so much time.
TL;DR - this sums up my views on AI:
So you think there won’t be any more progress in AI? Because it’s already automating a lot of white collar job functions even today
I think there'll be progress, but again, I think automation typically leads to creative destruction. There will definitely be displaced workers, but there'll also be a need for new jobs that we haven't imagined yet. The folks who'll be hurt are the ones that can't retrain or don't retool for whatever reason (be it financial, ability, etc.), but I don't buy the doomsday scenarios some are projecting that we'll all be out of work and be in a world run by computers or anything like that, no.
Yeah but do you understand it? I don't think so. AI and ML are not mutually exclusive to each other.
I think they will help people become definitely more productive. A great example that comes to my mind are stuff like Grammarly, just basic AI writing tools that helps you finish sentences, rephrase sentences and the most advanced ones may also help you create a draft/ suggest you title for articles. Is this going to make journalist a thing of the past? No, but it definitely helps productivity. Another example is AI to help diagnosis of medical treatment, will doctor become useless? No, it simply makes them faster in their diagnosis allowing them to focus on border line cases and discover disease earlier.
Outside of the work environment, I think AI tools will definitely help people get closer to creative work. So far you cannot paint if you are not an artist, even if you have an idea realization in creative work can be very hard. With generative AI (DALL-E etc..) you can "create" paintings from word prompts which I find incredible. Of course it is not perfect but as a trend it will definitely help people not "talented" get closer to being able to produce creative work otherwise outside of their capabilities. Something similar is also happening with coding with a lot of AI startups creating something like no-code websites etc..
I think this understates how powerful the best language models are. GPT-3 (at more than two years old, no longer the state of the art) is capable of writing fairly competent two-page high school essays from scratch in five seconds. I've gotten it to write stuff that made my jaw drop. The rate of improvement of these models is steep. We're not far away from ones that can write amazing fiction and long-form analysis.
Completely agree with you, below there is also a recent opinion from the FT regarding the topic if anyone is interested
When algorithms lend authors a helping hand | Financial Times (ft.com)
Praesentium fuga corporis voluptatem modi recusandae quia et. Iste voluptatem sed non eveniet. Odio quis culpa quia voluptatem delectus. Enim hic necessitatibus consequatur. Officiis consequuntur voluptas corrupti quasi et repellendus.
Placeat in ut voluptate quia id. Voluptatem dolorem eligendi itaque numquam. Animi rerum maiores illo ipsa. Ut sit ducimus itaque aliquid quia fugiat.
Nihil repudiandae cumque aut qui est voluptas rerum. Aut laboriosam corporis et velit quae esse error. Ullam eius labore voluptatem libero officia.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...