I am once again surfacing my life hack:
Traveling so much for work can put a strain on my wife. She’s juggling 3 kids, school activities, work, 3 dogs, and a bunch more.
I get lots of hotel travel points.
When I get back, I book my wife 1-2 days at a hotel down the street so she can read, relax, destress, or whatever she wants.
We started doing these mental health getaways during Covid and have kept up with them ever since!
Hola guys!
Amazon taught me one thing well - to build mechanisms.
In the last 3 years, I built 15+ GenAI projects as a head of ai in a consulting company serving to Fortune 500 companies.
I nailed one thing well - how to provide for LLMs a feedback loop.
Now my Claude code agents self-reiterate and self-perfect on projects with code and without.
Follow me to learn how to do it.
Just got done doing a virtual workshop for @MLHacks where I showed off @auth0 beyond just signin/signout.
Really proud of how this turned out! Amazing time and love teaming up with those folks!
Shoutout to Mr. AI Tech for coming in clutch with the praise 🤣
OAuth 2.1 is officially baking security best practices into the spec. From mandatory PKCE to the death of the Implicit Flow, here is what you need to know in 60 seconds.
My team at @auth0 has been shipping like crazy lately! This time we're super excited to announce our official integration with @ClaudeDevs.
By giving Claude Code knowledge of Auth0 best practices, you spend less tokens tweaking configs and more time delighting your users!
https://t.co/kcGWVGMOih
It's here and it's amazing! The team at @auth0 has partnered with @stripe to make it easier for developers to have a centralized workflow for configuring and managing their applications!
With the introduction of Stripe Projects (developer preview), your auth setup now lives alongside the rest of your application stack!
Learn more in our release blog! https://t.co/X2du73VVF3
The Authorized to Act Hackathon is heating up 🚀
Registrations are growing fast and builders are already shipping some impressive submissions. If you're building agentic AI, autonomous systems, or tool-calling apps, this is your moment.
Build secure AI agents that can act on behalf of users.
Register here → https://t.co/DRe33JBhDx #AuthorizedToAct
I am looking for one fall'24 intern, along the direction of LLM+reasoning/planning. The intern shall start before Nov this year. If you have interest, please contact [email protected]. Thanks!
I plan on keeping a close eye these next couple years on these 9 companies that could significantly impact the future of AI.
Vaire Computing (Reversible compute)
Liquid AI
Symbolica AI
Sakana AI
Kyutai
Holistic (France/No site yet)
Extropic
Normal computing
Rain AI
@ldjconfirmed I was looking up Symbolica and Liquid, which led me to your post. Those two companies absolutely intrigue me, so I'm excited to learn more about the others. Reversible compute makes me think of quantum, looking forward to seeing what they're about. thx.
I think AI agentic workflows will drive massive AI progress this year — perhaps even more than the next generation of foundation models. This is an important trend, and I urge everyone who works in AI to pay attention to it.
Today, we mostly use LLMs in zero-shot mode, prompting a model to generate final output token by token without revising its work. This is akin to asking someone to compose an essay from start to finish, typing straight through with no backspacing allowed, and expecting a high-quality result. Despite the difficulty, LLMs do amazingly well at this task!
With an agentic workflow, however, we can ask the LLM to iterate over a document many times. For example, it might take a sequence of steps such as:
- Plan an outline.
- Decide what, if any, web searches are needed to gather more information.
- Write a first draft.
- Read over the first draft to spot unjustified arguments or extraneous information.
- Revise the draft taking into account any weaknesses spotted.
- And so on.
This iterative process is critical for most human writers to write good text. With AI, such an iterative workflow yields much better results than writing in a single pass.
Devin’s splashy demo recently received a lot of social media buzz. My team has been closely following the evolution of AI that writes code. We analyzed results from a number of research teams, focusing on an algorithm’s ability to do well on the widely used HumanEval coding benchmark. You can see our findings in the diagram below.
GPT-3.5 (zero shot) was 48.1% correct. GPT-4 (zero shot) does better at 67.0%. However, the improvement from GPT-3.5 to GPT-4 is dwarfed by incorporating an iterative agent workflow. Indeed, wrapped in an agent loop, GPT-3.5 achieves up to 95.1%.
Open source agent tools and the academic literature on agents are proliferating, making this an exciting time but also a confusing one. To help put this work into perspective, I’d like to share a framework for categorizing design patterns for building agents. My team AI Fund is successfully using these patterns in many applications, and I hope you find them useful.
- Reflection: The LLM examines its own work to come up with ways to improve it.
- Tool use: The LLM is given tools such as web search, code execution, or any other function to help it gather information, take action, or process data.
- Planning: The LLM comes up with, and executes, a multistep plan to achieve a goal (for example, writing an outline for an essay, then doing online research, then writing a draft, and so on).
- Multi-agent collaboration: More than one AI agent work together, splitting up tasks and discussing and debating ideas, to come up with better solutions than a single agent would.
I’ll elaborate on these design patterns and offer suggested readings for each next week.
[Original text: https://t.co/y4McIAjD2m]