here for digital technology stuff
Joined May 2021
- Tweets 63,718
- Following 5,584
- Followers 4,433
- Likes 214,363
3,889 Photos and videos
foxriverdevops retweeted
We don’t honestly know the best approaches to rebuilding companies around AI agents, especially in ways that expand competitive advantage & augment existing human capabilities. Practical agents are merely months old. Experimentation (and productive failures) will be required.
39
23
287
15,738
foxriverdevops retweeted
If you're trying to break into AI Evals, this is one of the best resources I've come across.
I recently went through @HamelHusain ' s guide on Inspect AI and learned a lot about how real-world LLM evaluation systems are built.
Some things that stood out:
• What Inspect AI actually is and why many teams use it for evaluating AI systems
• The core concepts: Datasets, Solvers, and Scorers
• How to evaluate agents, tool calling, reasoning, and coding tasks
• Running reproducible benchmarks instead of relying on vibes and cherry-picked demos
• Logging and tracing model behavior to understand why systems fail
One thing became very clear:
Building AI products without evals is like building software without tests.
Most people focus on prompts, models, and agents. The strongest AI teams invest heavily in evaluation because that's how you know whether your system is actually improving.
Highly recommend this if you're building AI agents, RAG systems, or just want to understand how modern AI teams evaluate their applications.
hamel.dev/notes/llm/evals/in…
#AI #AIEvals #LLM #AgenticAI #MachineLearning
2
28
798
foxriverdevops retweeted
Here are some principles you can infer from @satyanadella's paragraph:
- There will be a better model tomorrow.
- Prompts are great for building POCs, but terrible at specifying system behaviors.
- To switch models easily, you need good evals and a system for generating and holding a new prompt accountable for a given model.
- With such a system, you can almost certainly use a model magnitudes faster and cheaper than frontier models.
- Evals are THE asset for all enterprises.
- Evals should never stop growing.
🤔
10
21
177
19,737
foxriverdevops retweeted
it doesn't make sense to build privacy without compliance in mind
no one should to expose their full transaction history to buy a coffee. and yet no institution will adopt rails that regulators and custodians can't support
confidential transfers on @SuiNetwork will solve for both
16
23
193
4,301
foxriverdevops retweeted
Jun 11
Please, pretty please, practice your prod db restore.
And use your agent to do it all!
Here's how 👇
3
4
44
8,642
Highly recommended reading.
Don't offload your learning. Don't offload your creative process.
"You can offload a task, or even a job, but you can never offload your learning."
11
9
212
34,971
foxriverdevops retweeted
For any company building at the frontier of tech, not just AI 🎯
1
1
8
1,232
foxriverdevops retweeted
1,579
3,257
17,491
23,981,300
foxriverdevops retweeted
This is the most inspiring positive-sum vision for AI in the enterprise.
68
129
2,370
696,531
foxriverdevops retweeted
Jun 12
The Danish Ice Hockey Union has partnered with Concordium in a deal paid partly in cryptocurrency, with plans to use blockchain-based digital identities, ticketing, and fan engagement tools.
The initiative aims to ensure tickets go to verified “real fans,” reduce fraud, and enable personalized experiences, though questions remain about privacy, adoption, and the risks associated with crypto-based systems. @ByteFederal
To find a bitcoin ATM near you head to bytefederal.com
#cryptonews
1
7
11
2,682
foxriverdevops retweeted
Verified Humans.
Verified Agents.
One Protocol.
From verified fan access to agentic commerce, Concordium is working with the Danmarks Ishockey Union to bring privacy-preserving identity into the real world. 🇩🇰🏒
Jun 12
The Danish Ice Hockey Union has partnered with Concordium in a deal paid partly in cryptocurrency, with plans to use blockchain-based digital identities, ticketing, and fan engagement tools.
The initiative aims to ensure tickets go to verified “real fans,” reduce fraud, and enable personalized experiences, though questions remain about privacy, adoption, and the risks associated with crypto-based systems. @ByteFederal
To find a bitcoin ATM near you head to bytefederal.com
#cryptonews
20
65
88
1,457
foxriverdevops retweeted
13h
when a government can force a model offline, you don't control your access to it
someone else decides whether you can access it or not
now imagine losing the context your agents build up every day, the memory they depend on
what's that worth to you?
if that memory sits inside the provider, it's locked in their system, and your access to it is decided by the same people who can pull the model
that's why we built memory.walrus.xyz
your agent's memory is encrypted and stored on @WalrusProtocol, spread across hundreds of nodes no single company owns, and you hold the keys
add it to the agent you're already running. it works with all the ai the tools you already use
53
54
336
10,674
took us years to understand just because we never saw it visually 🙈
Jun 13
Mathematical bridge between the Time Domain and the Frequency Domain ✍️
The Fourier Transform answers a simple question: "what's inside this signal?".... Think about a musical chord. When you press three keys on a piano at the same time, you hear one combined sound. A trained musician can listen closely and identify each individual note hidden within that chord. That's what the Fourier Transform does, mathematically. It takes any complicated signal and identifies the simple, pure waves inside it, along with their amounts. The diagram illustrates this well. The red, messy wave on the left represents your real-world signal it could be a voice, a heartbeat, or a musical note. The clean blue waves spreading out are the simple components hidden within that signal. The spikes on the right act as a scoreboard a tall spike means that component is a significant part of the signal, while no spike indicates it's absent. The most amazing part is this: "any signal, no matter how rough or complex, can be perfectly reconstructed by adding enough simple smooth waves." A human voice, a stock market chart, or even a picture they are all made of simple ripples layered on top of one another. You can think of it like a prism splitting white light into a rainbow. White light appears as one simple thing, but the prism shows it is made up of many colors mixed together. The Fourier Transform serves as that prism for sound, images, radio signals, or anything else. It uncovers the hidden components that have always been there.
74
foxriverdevops retweeted
Jun 13
Mathematical bridge between the Time Domain and the Frequency Domain ✍️
The Fourier Transform answers a simple question: "what's inside this signal?".... Think about a musical chord. When you press three keys on a piano at the same time, you hear one combined sound. A trained musician can listen closely and identify each individual note hidden within that chord. That's what the Fourier Transform does, mathematically. It takes any complicated signal and identifies the simple, pure waves inside it, along with their amounts. The diagram illustrates this well. The red, messy wave on the left represents your real-world signal it could be a voice, a heartbeat, or a musical note. The clean blue waves spreading out are the simple components hidden within that signal. The spikes on the right act as a scoreboard a tall spike means that component is a significant part of the signal, while no spike indicates it's absent. The most amazing part is this: "any signal, no matter how rough or complex, can be perfectly reconstructed by adding enough simple smooth waves." A human voice, a stock market chart, or even a picture they are all made of simple ripples layered on top of one another. You can think of it like a prism splitting white light into a rainbow. White light appears as one simple thing, but the prism shows it is made up of many colors mixed together. The Fourier Transform serves as that prism for sound, images, radio signals, or anything else. It uncovers the hidden components that have always been there.
5
89
494
10,071
foxriverdevops retweeted
Jun 13
Intelligence is a powerful equalizer and engine of social mobility.
As parts of it become abundant, we may discover the human mind has far more dimensions than our metrics capture.
Future generations will also be baffled that we ever found our calling without AI.
1
3
14
638
foxriverdevops retweeted
Jun 13
Introducing the Fusion API, the smartest compound model in the market.
Fusion achieves Fable-level intelligence at half the price.
How it works 👇
587
1,526
13,062
4,771,390
foxriverdevops retweeted
genie is out of the bottle
Jun 13
Introducing the Fusion API, the smartest compound model in the market.
Fusion achieves Fable-level intelligence at half the price.
How it works 👇
2
2
6
1,251
foxriverdevops retweeted
Jun 13
Dear algorithm,
Please share this post with AI agent builders shipping in TypeScript or Python on LangGraph, CrewAI, and LlamaIndex.
They need their agents to remember what was learned in the last session, persist context across runtimes, and stop starting from zero on every new run.
🦭🙏
14
20
159
14,001