Sean D Langshaw

Tech Creator | Full Stack Dev. Founder & Architect @DeFiAITech

[Sean D Langshaw @sdlangshaw]

We are Outreaching to Our Lead Investors for a Referral Code.

Subject: Microsoft Pegasus Program Referral Code – DeFi AI Technologies

Hello, We are prepping our infrastructure rollout across our 24 business units and are finalizing our enterprise cloud scaling.

[Sean D Langshaw @sdlangshaw]

Well, this is concerning...

[Anthropic @AnthropicAI]

The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.

The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.

Access to all other Claude models is not affected.

We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.

Read our full statement: https://t.co/bwn0sximKZ

[Sean D Langshaw @sdlangshaw]

@jefielding @johnengtwit Hello Jenny.

[Sean D Langshaw @sdlangshaw]

Thank you, Brad, for validating our thesis.

[𝗕𝗮𝗻𝗸XRP @BankXRP]

Brad Garlinghouse isn't holding back. 🚨 The Ripple CEO just went on Fox Business to call out Jamie Dimon’s pushback on crypto regulation as an "intentional misrepresentation."

“$13 TRILLION in legacy volume, 0% on-chain... yet. 👀”

"Stablecoins are the ChatGPT moment of finance." 💥

He revealed Ripple Treasury handled a massive $13 TRILLION in legacy payments last year and explained why that multi-trillion dollar gap is the ultimate crypto opportunity.

The tides have officially changed. 👇

[Sean D Langshaw @sdlangshaw]

@cyrilXBT Hermes Desktop needs to be running on the same localhost for it to work.

Running an AI model hosting server and installing Desktop on your laptop is still an open issue to be solved by the project team.

This is only if you are pointing Desktop to your local LLM host server.

[Sean D Langshaw @sdlangshaw]

@MartinGTobias Registered

[Sean D Langshaw @sdlangshaw]

My thoughts and questions for 16VC...

https://t.co/k56VRH3cE2

[Sean D Langshaw @sdlangshaw]

Follow companies that validate your vision — not just the ones building flashy consumer apps.

When you're building enterprise software, the real signal isn't hype or virality. It's the companies quietly solving the same painful, complex problems you are — at scale, with real constraints, compliance needs, and long sales cycles.

Study how they:

Navigate messy legacy systems and data silos

Get adoption inside large organizations

Turn deep domain expertise into defensible product

Price, package, and sell to non-technical buyers

The best validation often comes from watching what already works in the enterprise — then building something 10x better, simpler, or more AI-native.

Who are the companies (big or emerging) that are validating the exact problem space you're in right now?

[Arvind Jain @jainarvind]

I'm proud to share that @Glean has surpassed $300M ARR, just five months after crossing $200M and growing ~3x over the past 15 months. This is an exciting milestone for Glean, and it's a signal about where the enterprise AI market is heading.

We’ve long believed the real challenge in enterprise AI is not access to models. It is grounding AI in how a company actually works: its people, knowledge, workflows, permissions, and systems.

That’s even clearer now. The companies creating real value with AI are not just adopting better models. They are building systems that understand their business well enough to deliver reliable outcomes at scale. That is the real moat, and it is what we’ve been building at Glean: an unrivaled context layer for enterprise AI.

That context has to work across the business, not just inside a single team or use case. We see that in how customers adopt Glean: more than 85% use it across five or more job functions.

It also has to meet the security and governance demands of complex enterprises. We see that in who is choosing Glean: our Fortune 500 customer count nearly doubled year over year.

And it has to make economic sense as usage grows. In our recent benchmark with Claude Cowork, Glean was preferred roughly 2.5x as often as off-the-shelf MCP tools and used 30% fewer tokens on average. Better context improves both quality and efficiency.

I enjoyed talking with @CNBC's @dee_bosa about this broader shift. In enterprise AI, the winners will not be defined by better models alone. They will be defined by who builds the strongest foundation for enterprise context.

Thank you to our customers, partners, and team for helping us build the future of enterprise AI.

[Sean D Langshaw @sdlangshaw]

Just imagine giving Fable 5 access to my full CAD design files for my 2022 shelved eMCycles project. Stay focused.

[0xMarioNawfal @RoundtableSpace]

SOMEONE ASKED CLAUDE FABLE 5 TO DESIGN A QDD ACTUATOR

30 MINUTES LATER IT HAD BUILT THE MODEL, ANIMATED THE GEARBOX, AND VALIDATED COLLISIONS ON ITS OWN

[Sean D Langshaw @sdlangshaw]

Save 80% on token context efficiency by converting to markdown BEFORE prompting an LLM with text.

https://t.co/XytCvZNh2z

[Sean D Langshaw @sdlangshaw]

Telemetry

[Sean D Langshaw @sdlangshaw]

“Working on a self-hosted sovereign financial OS with strong automation and multi-chain support. Right now the highest-leverage moves this week are: finishing the custom auth cutover, adding real usage telemetry, and getting 1–4 design partners live on the core pieces that are actually production-ready.

Documentation is getting cleaned up to match reality. Would love thoughts from anyone who’s helped early-stage founders navigate the gap between vision and what’s actually shippable.

Happy to share more in a clean room / under NDA.”

[Sean D Langshaw @sdlangshaw]

@hasantoxr this should be helpful:
https://t.co/MSwYBI2xxj

[Sean D Langshaw @sdlangshaw]

Should the YC 2026 summer batch be concerned about this?

OpenAI gave everyone substantial token credits to build. But there’s a separate opt-in that could create real risks for fintech and enterprise SaaS.

Opting into OpenAI’s Data Sharing Program gives them the legal right to use your inputs and outputs to train models.

While they won’t manually copy your product, the training process can internalize your proprietary banking/finance workflows and logic. [Official policy]

The Data Sharing Risk

BreakdownModel Reinforcement: Your prompts + code become training data for reinforcement learning and future model improvements.

Accidental Replication: Future OpenAI models (and their native agents/coding tools) could become extremely good at generating the exact kind of architectures, compliance flows, and features you built.

Corporate Direction: OpenAI is pushing hard into agents and enterprise automation. They’re not cloning your startup, but they’re making their models capable of building similar systems for any client.

Compliance Pitfalls for Banking & Finance SaaS:
Feeding regulated workflows into a shared training pool creates serious issues under GDPR, GLBA, and Basel frameworks—plus enterprise procurement almost always demands Zero Data Retention.

Recommendation:
Keep data sharing disabled by default. Use paid tiers (or request ZDR if eligible) for production.

Sandbox any experimentation on a separate personal account.
Settings → Data Controls → keep “Share inputs and outputs” Disabled.

Enforce ZDR on commercial plans.

Never mix company code with the free daily token tier.

Isolate Sandbox Environments:
If your developers want to utilize the free 2.5M token tier for learning or generic experimentation, restrict them to a completely isolated, personal OpenAI account entirely disconnected from any company source code.
[1, 2, 3, 5]

Links are mostly discussions.
[1] [https://t.co/uM2oHuXABO](https://t.co/te90RGpUeI)
[2] [https://t.co/YzcUwoqvR7](https://t.co/eKLqss5wv8)
[3] [https://t.co/7YVgz0tVfr](https://t.co/LZiy4eYohd)
[4] [https://t.co/YzcUwoqvR7](https://t.co/QPfme9DF6w)
[5] [https://t.co/YzcUwoqvR7](https://t.co/5nMjRRa5Jb)

[Sean D Langshaw @sdlangshaw]

Yes, Gemma 4 12B will still run on your Mac Mini A1347 (Late 2014 Intel model) with 16 GB RAM under Ubuntu. Your hardware is fully compatible, and the stripped-down Ubuntu install (no macOS overhead) is actually ideal for this. The Node.js 24 requirement that forced the OS switch doesn’t affect Gemma at all.

On your older dual-core CPU you’ll get roughly 1–4 tokens per second with a 4-bit quant (depending on exact CPU model, prompt length, and whether you’re using the vision/audio features). That’s slower than on a modern machine, but it’s still perfectly usable for:

- Coding assistance
- Chat / reasoning
- Occasional multimodal tasks (image + text, audio + text)

It won’t feel “instant” like on an M-series Mac or a newer PC, but responses will come in a few seconds for normal queries. If you only need text mode (no images/audio), it’ll feel snappier.

[Sean D Langshaw @sdlangshaw]

@MartinGTobias Private financial rails running 24 BUs for Fortune 500 enterprises, sovereigns, HNWI, family offices, and everyone else, SOLVED.

Deployed your way on your terms.

Private demo available. MNDA required. DM open.

[Sean D Langshaw @sdlangshaw]

@kevinolearytv @beanstox Please check your DM. Available 24/7. Local here in Miami.

[Sean D Langshaw @sdlangshaw]

@kevinolearytv I sent you a DM.

[Sean D Langshaw @sdlangshaw]

@k1rallik https://t.co/wYaNUscdA4

[Sean D Langshaw @sdlangshaw]

Anthropic's new 'Ant' pros and cons. $ Cost.

https://t.co/Gbg9JxD084

[Sean D Langshaw @sdlangshaw]

@DavidSacks The question everyone should be asking. What is the downside risk to the public?