OpenAI's API & Function Update, Microsoft Guidance, Andromeda Cluster, Meta Open Source | E19
The future is rushing at us like a freight train - will AI crush our dreams by 2043 or give us indefinite life extension? Explore the tangled web of AI hype, loneliness, and alcoholism with your hosts as they dive deep into OpenAI's latest updates, Zuck's vision for chatbots gone wild, and whether Big Tech's sudden pivot to "openness" is too little too late. From Microsoft's "Guidance" to steering GPT-3.5, fall down the rabbit hole of what's really going on behind the scenes—where are we headed and how soon will robots replace that guy in accounts payable (spoiler: probably not as soon as he thinks!). An epic voyage to the outer reaches of speculation that will leave you equal parts informed, entertained and possibly needing a stiff drink.Try MusicGen here: https://huggingface.co/spaces/facebook/MusicGenIf you like this podcast please considering subscribing to the channel, liking and leaving a comment. We appreciate your support.Chapters---00:00 - Mark Zuckerberg's Redemption Arc00:14 - OpenAI's Function Calling, API updates and 16k GPT3.510:20 - Microsoft Guidance18:05 - Andromeda Cluster for startups26:24 - France's Mistral AI raises $113M seed round to take on OpenAI33:40 - Mark Zuckerberg's comments about open source on Lex Friedman40:27 - Deciding which LLM to use and when44:13 - MusicGen by Meta & Adobe Firefly48:14 - The Beatles are releasing a new song thanks to AI53:37 - "Transformative AGI by 2043 less than 1% likely" paper1:03:45 - Is AI Making Us Sad Drunks?Sources:---https://openai.com/blog/function-calling-and-other-api-updateshttps://github.com/microsoft/guidancehttps://andromedacluster.com/https://techcrunch.com/2023/06/13/frances-mistral-ai-blows-in-with-a-113m-seed-round-at-a-260m-valuation-to-take-on-openai/?tpcc=tcplustwitterhttps://twitter.com/polygonmojo/status/1667183372525789185?s=20https://huggingface.co/spaces/facebook/MusicGenhttps://www.bbc.com/news/entertainment-arts-65881813https://vulcan.io/blog/ai-hallucinations-package-riskhttps://gizmodo.com/chatgpt-detector-ai-kansas-research-paper-99-accuracy-1850519081https://www.theregister.com/2023/06/13/ai_loneliness_alcohol_study/https://forum.effectivealtruism.org/topics/open-philanthropy-ai-worldviews-contest#:~:text=The%20goal%20of%20the%20contest,AI%20timelines%20and%20AI%20risk.https://arxiv.org/abs/2306.02519