5 years from now, which company would be leading in the AI race? OpenAI seems to have started it but will it be the market leader even 5 years now? Remember Google wasn’t the first search engine; Facebook wasn’t the first social network; #openai #anthropic #cohere #mistral #google #meta #microsoft #amazon #ai #machinelearning
Microsoft will own OpenAI in 5 years on paper instead of how they own them now off the books
Where is XAi? Didn’t they just get a 20B dollar funding round?
In 5 years it will be clear that AI is a scam.
Apple
Based on Siri? Based on Titan ? These two are high profile AI projects they been working on for more than a decade
My guess is none and all of them will still be trying to make it work. Gen AI is still too unreliable and expensive to run. It will need more than 5 years maybe 10 to become economically sensible and useful. I’m working on a team that uses gen AI extensively both internally and customer facing, it has a lot of issues.
What are these internal and customer facing use cases ? Cos there are many productivity related use cases where it seems to work.
@yelp They do work, just expensive and unreliable. We use it to extract information from webpages. Basically input html and output data in the page. Expensive is an issue we all know, reliability is an issue because the information return can be wrong or sometimes not in the format we want even with prompt engineering. Same issue for external use. Expensive and sometimes just straight up lies
Google. 1. Very difficult to change consumer behavior and ain't no one going to bing.com . Google's models will be 90% as good as those developed by smaller companies which will benefit from greater agility compared to Google. 2. Google will attempt (note, not fully realize) to constrain models to not hallucinate and continue to artificially make them woke. 3. In the absence of any breakthroughs in hardware, Google's TPUs will be most efficient in model serving leading to better financials. And of course the Gen AI bubble will burst sometime in the next 5 years when ordinary people realize this is not AGI.
Some company not on the list that will be unafraid to dump the current crufty AI stack and replace it with something much simpler and more robust.
5 years from now better sense will hopefully prevail that AI is a sham
the top two will be a closed fully vertical platform vs open source. 1. Closed: Apple/Google 2. Open: Llama/bloom? Hard to imagine smaller players keeping up with the insane infra and data costs. 50/50 OpenAI open sources to keep up.
Who cares though? I’ve worked in AI/ML for almost a decade and the amount of non-expert postulating is innacurate and ultimately pointless.
this. OP means the cartoonist generative version aka bs you cant rely one. Good for some application though: code, porn, resumes etc