If you had to bet in one "AI startup" for the next 10 years, which one would it be and why?
hmm, Apple is the likeliest as far a single-firm AI products/services I still think the most ambitious marketplace/protocol wins over everything else though.
Lots of AI startups won't make it to 10 years. Is their half-life even greater than 10 months? https://open.spotify.com/episode/6O0rmMyZQDAUtYEdSEMrEk?si=8ab2e4202a054f70
Perplexity Find it genuinely helpful Will try making it the default search engine Not yet good for real time stuff though
Whoever truly nails a seamless realtime consumer-friendly gpt hardware device. @kepano predicted apple will add camera to earbuds w gpt-like model I want to walk around w minimal hw, no phone/glasses, inside/outside w full access to gpt*music and want the most minimal piece of hw, so ye...likely earbuds.
Character AI like @kn said, their models are better at role playing a specific character and I can see this becoming more of an bot-human social network than just chat Also open-source models like hugging face, civitai
https://www.neurophos.com — Either bet on low-level (Teams building hardware that makes training/inference faster) or bet on high-level (Teams providing the solved customer experience with private data sets and expert labelers). Wrappers are a path to acquihires, not 10x+ ROI. This is a low-level bet.
Stating the obvious but openai. It’s a wide open market but they’re way ahead and keep bringing out great stuff. So I’d bet on them. If they’re still considered a startup.
Hugging Face; they seem most likely to be an IETF or Wikipedia like org for the totality of AI. Meanwhile, Perplexity is my daily AI ‘device’; like if they made a phone I would toss mine and buy it.
Whoever most effectively embodies AI. Could be Tesla, could be Jony Ive’s collaboration with OpenAI, or something currently in stealth.
Perplexity seems like a good bet once they can can go real time and use the most capable models as default