Tiktok for text is here

Hey folks, this is Ben's Bites - your AI bestie.

Yesterday I found out Veal Milanese is actually from Milan...And that it tastes the same pretty much anywhere. We had our first meal out last night and now getting ready to go shopping all day as it's my wife's birthday - so happy birthday to her! (no she doesn't read this or know remotely what I do for a living)

Let's get to it, I hope your day is less expensive than mine.

Have a product, service, job, event, newsletter, app, book, movie, tool, or anything you'd like to share with over 33k subscribers? 

Advertise with us | Job board | Join Discord

🤌 Ben's Picks

1/

Massive funding round for storytelling start-up Tome

That is one monster funding round - $43m, valuing the start-up at a whopping $300m, before revenue. When you hear of Tome as a “storytelling” company, you’re probably imagining less than it is - but these guys are lining up to be PowerPoint’s first big competitor. By “story” they mean “narrative”, and the idea is AI-generated tiles able to create a compelling narrative behind anything: children’s stories, pitch decks, 3D prototypes - you name it. (Basically PowerPoint but… better.) This is a really cool start-up and provides some welcome competition. But what’s Microsoft’s response going to be when they get backed into an AI corner?

See an unpaywalled Forbes article here. And from the investors: Why Lightspeed is leading a $43M Series B in Tome.

2/

Personalised news with Artifact app made available to all

The Instagram co-founders have put together a team of pure tech veterans to muster up the newest news service on the block. It’s called Artifact, it centres itself around personalisation and social networks, and is now available to everyone. If you’re like me, you’re asking yourself: “Aren’t all news services personalised these days? What’s new?” That’s true, but they’re relying on a few core aspects making them strong competition in a highly saturated market - and the biggest one is the presence of social/interest bubbles. Your suggestions will be influenced by what your in-app contacts are reading. I’m interested…

Medium publication here.

3/

Spotify to launch an assistant DJ AI whilst you listen to music

I actually think that Spotify’s music suggestions are spot on (pun definitely not intended), so I’m pretty intrigued by this new tool that they’re set to announce at their upcoming Stream On event. It’s called ‘DJ’, and as far as I can tell it’s a combination of music suggestions, synthetic voice generation, and generative chat to let you join an actual DJ-type-AI whilst you listen. It’ll introduce songs, provide some background, and guide you through your tunes. It seems pretty cool, and looks to be more interactive than something that just moves through music suggestions, hopefully allowing it to learn your preferences even faster.

4/

Get app and website mockups from a prompt with Autodesign by Uizard

I’m seriously inept when it comes to design creativity, so when I see AI getting applied to create tools that’ll help me design creatively it always piques my interest. Uizard (yeah, awesome name) have presented Autodesign, their new tool that allows you to enter a prompt and get a load of multi-screen mock-ups for UIs on apps and websites. The demo looks pretty sweet, and something I’d definitely use. Man, I can’t stop thinking about how cool their name is!

🛠️ Cool Tools

• Zapier + Leap AI integration - adding AI to your app just got 10x easier. (link)

• Landing AI from Coursera’s co-founder, releases computer vision tool for anyone to use. (link)

• InvoicesGPT - Fully automate the handling of all incoming invoices. (link)

• Fathom - Upgrade your productivity with your own free AI meeting assistant. (link)

• Ad Creative - Generate conversion focused ad creatives and social media post creatives in a matter of seconds. (link)

• AskHN - The collective GPT-embodied wisdom of Hacker News. (link)

• Raycast AI - Write smarter, code faster and answer questions quicker. (link)

• Notion AI is now available to everyone. (link)

• Chart - Package your models into high-performant C++ servers and deploy them into your own cloud account. (link)

• Byterat - Data platform for battery engineers that unlocks the full potential of battery testing data. (link)

• Lavo Life Science - AI-Accelerated quantum chemistry for rapid drug development. (link)

• SHP - Stanford human preferences dataset of 385k collective human preferences over text. (link)

• Runway Studios - A production partner and platform for the next generation of storytellers. (link)

• Artifact - The tiktok for news, now open to everyone. (link)

🤓 Miscellaneous

• Emergent capabilities of ML models - Deception and optimization. (link)

• China regulators rein in AI chatbots over fears of uncensored replies. (link)

• Chatbots don’t have feelings, but people do. We need to start thinking about the consequences. (link)

• Samsung announces enhancement of Bixby. (link)

• Introducing Cohere summarise beta - A new endpoint for text summarization. (link)

• A new era of storytelling at Runway ML’s first annual AI film festival. (link)

• Google’s progress toward quantum error correction. (link)

• JPMorgan restricts staff from using ChatGPT. (link)

• I made an AI clone of myself. (link)

• Ethics & society at Hugging Face. (link)

• McKinsey & Company joins corporate affiliate program at Stanford HAI. (link)

• Future of AI in VC. (link)

• OpenAI’s Foundry will let customers buy dedicated compute to run its AI models. (link)

• Spoke AI raises $2.1M for using generative AI to pull signal from workplace noise. (link)

• China’s first ChatGPT like chatbot MOSS released for public testing. (link)

🎓 Learn

• Complete guide to ControlNet for Stable Diffusion img2img. (link)

• Different parallelism approaches to train models like ChatGPT and stable-diffusion? (link)

• Introduction to PPO with sample-factory. (link)

🔬 Research

• Fair Diffusion - Instructing text-to-image generation models on fairness. (link)

• SeMani - Semantic manipulation for entity-level text-guided image manipulation. (link)

• Scaling robot learning with semantically imagined experience from Google. (link)

• Reduce, Reuse, Recycle from Google - Compositional generation with energy-based diffusion models and MCMC. (link)

• Guiding large language models via directional stimulus prompting from Microsoft. (link)

• How does in-context learning help prompt tuning, from Microsoft. (link)

• Modular deep learning from Google. (link)

• Edgeformers - Graph-empowered transformers for representation learning on textual-edge networks. (link)

• KS-DETR from Toyota - Knowledge sharing in attention learning for detection transformers. (link)

• Connecting vision and language with video localised narratives from Google. (link)

• Vid2Avatar - 3D avatar reconstruction from videos in the wild via self-supervised scene decomposition. (link)

• Conformers are all you need for visual speech recognition from Google. (link)

• UML from Google - A universal monolingual output layer for multilingual ASR. (link)

• Steerable equivariant representation learning. (link)

• RoboNinja - Learning an adaptive cutting policy for multi-material objects. (link)

📰 Unclassifieds

• Get a proven, tested blueprint for creating high-quality content using AI. (link)

• Newcomer and Volley are hosting an exclusive, one-day AI summit on March 30, 2023 in Cerebral Valley, SF. Apply to attend here >>

• Recollect is hiring senior software engineers. Apply here for backend and data.

• BlueCallom Corp. is hiring a sales manager in Zurich. Learn more >>

Have a product, service, job, event, newsletter, app, book, movie, tool, or anything you'd like to share with over 33k subscribers? 

Advertise with us | Job board | Join Discord

🖼 AI images of the day

🤗 Share Ben's Bites

Send this to 1 AI-curious friend and receive my AI project tracker database!

or copy/paste this link: https://bensbites.beehiiv.com/subscribe?ref=PLACEHOLDER

⭐️ How did we do?

⭐️ REAL REVIEWS