TheSequence - The ChatGPT Challengers
Was this email forwarded to you? Sign up here The ChatGPT ChallengersSundays, The Sequence Scope brings a summary of the most important research papers, technology releases and VC funding deals in the artificial intelligence space.📝 EditorialLast week, in this editorial, we presented the thesis that, with the release of ChatGPT and the upcoming GPT-4, OpenAI has hit “escape velocity” relative to the rest of the AI market. We try our best to keep these editorials short and not annoying, so I didn’t have enough space to present the other side of the argument related to the most imminent competition to OpenAI and ChatGPT and GPT-4. While ChatGPT has become the most popular release in the history of AI and GPT-4 is already positioned to get massive distribution via Microsoft products, there are some interesting alternatives to keep an eye on. Most of these models are likely to be released in the next few months so we still don’t have a lot of empirical evidence of their capabilities and limitations, but they look promising on the surface. 1) Claude: Anthropic has become one of the hottest AI startups at the moment with the promise of Claude, a language agent that follows similar design practices as ChatGPT and GPT-4. Claude is still in very limited beta but Anthropic has managed to attract over $1 billion in venture funding. 2) LaMDA: Google’s anticipated ChatGPT alternative should become available at some point this year. LaMDA incorporates interesting ideas such as chain of thought reasoning that could enable novel capabilities relative to ChatGPT or GPT-4. 3) Sparrow: DeepMind also seems to be working on a ChatGPT alternative called Sparrow which should be released on private beta this year. Given DeepMind’s expertise in reinforcement learning can be a strong differentiator as these techniques have proven paramount in the creation of InstructGPT and ChatGPT. 4) Stability AI: The creators of Stable Diffusion have publicly said they are working on an open-source version of ChatGPT. There are interesting expectations around what AI powerhouses like Meta and startups such as Chatsonic will do in the space. For now, it seems that Anthropic and Google-DeepMind are closer to provide a viable alternative to ChatGPT. It is also interesting to think about the impact that Stability AI could have in the market with an open-source release similar to Stable Diffusion. Picking winners this early in the generative AI space might be a fool’s errand. The frantic pace of innovation might still allow new startups to emerge and compete at scale with the big AI labs. Not easy, but certainly possible with the amount of capital chasing this market. 🔎 ML ResearchInstruction Tuning and LLMsGoogle Research published a paper introducing FLAN, a collection of tasks, methods and templates for instruction tuning in large language models —> Read more. CausalAISalesforce Research published a paper detailing CausalAI, a library for determining causal relationships in time series and tabular data —> Read more. ML Systems and InfrastructureGoogle Research published a comprehensive summary of their research in large scale ML infrastructure, , hardware, programming and sustainability —> Read more. 🤖 Cool AI Tech ReleasesChatGPT PlusOpenAI launched ChatGPT, a commercial version of the popular chatbot that enables improved performance, reliability and access to new features —> Read more. AI Generated Text ClassifierOpenAI launched a classifier that can differentiate between human and AI generated text —> Read more. Computer Vision at Hugging FaceThe Hugging Face team discusses the current libraries and capabilities in their platform for computer vision models —> Read more. VizierGoogle Research open sourced Vizier, a framework for black box, large scale hyperparemter optimization —> Read more. Ray 2.2Ray, the popular ML scalability platform, unveiled its 2.2 release with new features around observability, high performance and reproducibility —> Read more. 🛠 Real World MLReal Time Predictions at LyftLyft discusses the architecture to serve models for real time predictions —> Read more. 💸 Money in AI
You’re on the free list for TheSequence Scope and TheSequence Chat. For the full experience, become a paying subscriber to TheSequence Edge. Trusted by thousands of subscribers from the leading AI labs and universities. |
Older messages
💡Share Your Thoughts on Applied ML for a $25 Amazon Gift Card*
Friday, February 3, 2023
As a member of the ML community, we'd love for you to participate in our industry survey—it'll only take 10 minutes, and the first 150 respondents will receive a $25 Amazon gift card! Your
Edge 266: The Magic Behind ChatGPT: Reinforcement Learning with Human Feedback
Thursday, February 2, 2023
One of the techniques that enable the ChatGPT breakthrough comes from a 2017 research paper.
📍 Free Guide: Maximize the ROI of your AI/ML Investment: Building vs. Buying Monitoring Solutions*
Wednesday, February 1, 2023
There is no one-size-fits-all solution for ensuring model performance and accuracy
Edge 265: Interpretability Methods for Deep Neural Networks
Tuesday, January 31, 2023
Interpretability methods optimized for deep neural networks, OpenAI's interpretability technique to discover multimodal neurons on CLIP and the Eli5 framework.
Has OpenAI Hit Escape Velocity?
Sunday, January 29, 2023
Sundays, The Sequence Scope brings a summary of the most important research papers, technology releases and VC funding deals in the artificial intelligence space.
You Might Also Like
Corporate Casserole 🥘
Monday, November 25, 2024
How marketing and lobbying inspired Thanksgiving traditions. Here's a version for your browser. Hunting for the end of the long tail • November 24, 2024 Hey all, Ernie here with a classic
WP Weekly 221 - Bluesky - WP Assets on CDN, Limit Font Subsets, ACF Pro Now
Monday, November 25, 2024
Read on Website WP Weekly 221 / Bluesky Have you joined Bluesky, like many other WordPress users, a new place for an online social presence? Also in this issue: CrawlWP, Asset Management Framework,
🤳🏻 We Need More High-End Small Phones — Linux Terminal Setup Tips
Sunday, November 24, 2024
Also: Why I Switched From Google Maps to Apple Maps, and More! How-To Geek Logo November 24, 2024 Did You Know Medieval moats didn't just protect castles from invaders approaching over land, but
JSK Daily for Nov 24, 2024
Sunday, November 24, 2024
JSK Daily for Nov 24, 2024 View this email in your browser A community curated daily e-mail of JavaScript news JavaScript Certification Black Friday Offer – Up to 54% Off! Certificates.dev, the trusted
OpenAI's turbulent early years - Sync #494
Sunday, November 24, 2024
Plus: Anthropic and xAI raise billions of dollars; can a fluffy robot replace a living pet; Chinese reasoning model DeepSeek R1; robot-dog runs full marathon; a $12000 surgery to change eye colour ͏ ͏
Daily Coding Problem: Problem #1618 [Easy]
Sunday, November 24, 2024
Daily Coding Problem Good morning! Here's your coding interview problem for today. This problem was asked by Zillow. Let's define a "sevenish" number to be one which is either a power
PD#602 How Netflix Built Self-Healing System to Survive Concurrency Bug
Sunday, November 24, 2024
CPUs were dying, the bug was temporarily un-fixable, and they had no viable path forward
RD#602 What are React Portals?
Sunday, November 24, 2024
A powerful feature that allows rendering components outside their parent component's DOM hierarchy
C#533 What's new in C# 13
Sunday, November 24, 2024
Params collections support, a new Lock type and others
⚙️ Smaller but deeper: Writer’s secret weapon to better AI
Sunday, November 24, 2024
November 24, 2024 | Read Online Ian Krietzberg Good morning. I sat down recently with Waseem Alshikh, the co-founder and CTO of enterprise AI firm Writer. Writer recently made waves with the release of