- Enzyme
- Posts
- bigger than the Internet đ
bigger than the Internet đ
your web3 & AI weekly roundup powered by Leah from OneProfile.ai.

is this a forward? join here.
gm gm,
welcome to one more week closer than ever to a world where we all just Sign in with ChatGPT.
Nvidia stole the show with a record $44B in revenue, up 69% and brushing off China concerns like theyâre background noise.
Meanwhile, Anthropicâs CEO warns AI might wipe out half of all entry-level white-collar jobs in just five years.
As always, weâre using this space to provide you with the latest product updates and relevant industry news in the shortest form possible.
Letâs go! @getenzyme
IN PARTNERSHIP WITH LEAH FROM ONEPROFILE
Still wasting time & $$$ to hire a dev?

Hiring great developers is a full-time job.
One most teams donât have time for.
Between sifting through low-signal applications, coordinating interviews, and chasing no-shows, itâs easy to burn weeks (and serious budget) before you even make an offer.
Thatâs where Leah steps in â a superhuman AI hiring agent built to find, screen, and hire world-class tech talent for you.
Leah sources globally, runs assessments, manages interviews, and surfaces only the top matches ready to deploy. Add her to Slack, and youâll cut your hiring costs by up to 80% instantly âď¸
Weâre opening up partnership opportunities for teams who want to start hiring right away with Leah on their side. If your teamâs looking to scale smarter, not harder đ
IN THE SPOTLIGHT
Claude 4 is here!
And it might be the best coding model yet.
Introducing the next generation: Claude Opus 4 and Claude Sonnet 4.
Claude Opus 4 is our most powerful model yet, and the worldâs best coding model.
Claude Sonnet 4 is a significant upgrade from its predecessor, delivering superior coding and reasoning.
â Anthropic (@AnthropicAI)
4:36 PM ⢠May 22, 2025
Anthropic just dropped Claude 4, their latest and most powerful AI model lineup, and itâs making waves especially among developers. The new release includes Claude Opus 4 and Claude Sonnet 4, both designed to handle complex reasoning, long-term memory, and agent-like tasks.
Opus 4, in particular, scored a record-breaking 72.5% on the SWE-bench coding benchmark, outperforming even OpenAIâs GPT-4.1. It can stay focused on large-scale software projects for up to seven hours straight, no memory wipes, no getting lost in the sauce.
One of the standout features is what Anthropic calls âhybrid reasoningâ. You get fast replies for quick interactions and deep-thinking mode for longer tasks like research, planning, and decision-making. These models can refactor legacy codebases, coordinate across teams, and even manage full-stack systems with minimal help. And they donât just forget what happened in the last session making the Claude 4âs long-term memory finally tackles the "AI amnesia" problem.
Both models are also wired for autonomy. They can break down goals, plan steps, and execute tasks like agents, even calling tools mid-conversation. In fact, during testing, Opus 4 played PokĂŠmon Red for 24 hours straight without derailing. Whether youâre building an internal code review agent or a long-running research assistant, Claude 4âs flexibility makes it production-ready. Sonnet 4 leans into speed and efficiency, while Opus 4 is the go-to for harder problems.
You can access both models through Amazon Bedrock, with enterprise-level integration and safety controls built in. And yes, Claude 4 models are trained up to March 2025 but behave as if their knowledge ends in January (thatâs a built-in safety feature).
Anthropic is clearly gunning for the top spot, and with Claude 4, they might have just claimed it for long-haul, high-stakes AI work.
SUBSTRATES
What else is on the radar?
WEB3
Circle files for NYSE IPO, marking a significant step for US crypto integration and stablecoins validation in public markets.
The soccer giant Paris Saint Germain adopts a Bitcoin treasury reserve.
MetaMask now supports Solana on their wallet.
Solanaâs Co-Founder Raj Gokalsâ data was leaked on Migosâ hacked Instagram account.
AI
DeepSeek releases R1 updated version with a permissive MIT license for commercial use on Hugging Face.
Perplexity releases Perplexity Labs, new feature for Pro subscribers that allows report generation, spreadsheets, dashboards, simple web apps and more with AI.
OpenAI upgrades Operator with o3, achieving state-of-the-art performance on multiple benchmarks.
Opera works on Neon, an AI browser that âcan act on your behalfâ by understanding whatâs happening on your screen and react accordingly.
Mistral launches Codestral Embed, new model that tackles programming tasks.
FROM THE REPOSITORY
Check this out!
All the AI hallucinations cases so far in one database.
The 100 things Google announced at I/O.
Use now or save for later:
Streamline and optimize your prompts with PromptGround.
BityClips lets you create stunning videos from a single prompt.
Discover âthe whyâ behind your customersâ behaviour with Stratify.
*Shipping amazing products or services in AI or web3? Just hit reply and share your story with our 1k+ communityâtotally free!
Visual of the week: The top chip designers by revenue
SCOOPS OF THE WEEK
|
|
And thatâs it for the week - not so short but sweet.
Keep on building!
P.S. Got a friend who needs this? Weâd appreciate it if you passed it on!
Reply