It's on! Google and MS go head to head while gen AI vals fly.

Googles Bard is coming as Microsoft rushes a press release to be the first to announce the gen ai and search merger.

Welcome to RiD! My side / passion project on all things language AI. It’s basically where you find out today, whats happening tomorrow (hopefully).

Attribution - elevenlabs.io

Here is the TL;DR list of this weeks newsletter:

• Google releases Bard and Microsoft / OpenAI go head to head.

• Valuations in generative AI are flying.

• Why we’re (probably) not all going to get replaced by robots.

• RLHF keeps rearing it’s (not so) ugly head

• Text to video is getting closer.

Google releases Bard and Microsoft / OpenAI go head to head

Google has brought forward, it would seem, the release of it’s own LLM called Bard. Given ChatGPT was the fastest growing app in human history, it’s not surprising they want to minimize as much as possible OpenAI & Microsoft’s first mover advantage. Google will unveil Bard later this week, which prompted Microsoft to then announce it’s own press conference Tuesday to talk all things search and ChatGPT. Take a look at the live updates for some insights into what was unveiled.

Valuations in generative AI are flying

More widely, the generative AI ride is getting wild and valuations are soaring. We saw a couple of large raises this week per below however other more “established” companies valuations are seeing a nice pop.

Pete @nonmayorpete

Valuations: OpenAI: $29 billion Anthropic: $5 billion Character AI: $1 billion Cohere: $6 billion We love the big numbers.

Why we’re (probably) not all going to get replaced by robots

A good piece discussing why we (probably) won’t enter a language AI hall of mirrors. Clearly various scenarios are possible and we wont live in a binary good/bad outcome scenario but it’s good to hear an optimistic take on the direction we’re heading.

All the way with RLHF(ey)

A theme in this newsletter is differentiation through RLHF. More on how it was leveraged for chatGPT.

Edge 266: The Magic Behind ChatGPT: Reinforcement Learning with Human Feedback

Here is some more input on this topic. This is a little self serving but still directionally correct.

Alexandr Wang @alexandr_wang

we're starting to see top companies spend the same amount on RLHF and compute in training ChatGPT-like LLMs for example, OpenAI hired >1000 devs to RLHF their code models crazy—but soon companies will start spending $ hundreds of Ms or $ billions on RLHF, just as w/compute

Text to video is getting closer

As I said last week, as of now AI / robots can:

• Read & write

• Talk - 11Labs (see above)

• Hear

• See

• Dance / navigate the physical world

Video augmentation however is still a WIP. This example though (floating around twitter recently) show how close we are getting. This still requires some degree of expertise and lift but pointing to where we are heading.

Zain Kahn @heykahn

Ok this is insane. @Flawlessai let's you change the dialogue on a video and the lips sync absolutely perfectly to each word.

Industrial scale deep fakes already have people scrambling.

Basic market data tracker:

Happenings this week:

Commercial:

• Ushur, which aims to automate aspects of the customer experience, raises $50M.

• Report: China’s Baidu is preparing a ChatGPT chatbot rival

Technical:

• Microsoft AI Research Proposes eXtensible Prompt (X-Prompt) for Prompting a Large Language Model (LLM) Beyond Natural Language (NL)

• Researchers at Stanford Introduce Parsel: An Artificial Intelligence AI Framework That Enables Automatic Implementation And Validation of Complex Algorithms With Code Large Language Models LLMs

Ethical & other

• As AI applications like ChatGPT and DALL-E evolve, we'll see even bigger concerns over issues of copyright and bias.

• AI cannot be credited as authors in papers, top academic journals rule.

»»»»»»»»»»»»»»»»»»>