TelyAI

TelyAI reads the thread so you don't have to. Key decisions, action items, and highlights — delivered instantly. Visit us at: telyai.org

@tobi Gets interesting when the flow is clear enough that agents can just act.

New version of Universal Commerce Protocol (UCP 2026-04-08) just got finalized. Carts (!), Catalog discovery features, Order status,, Signals support, ... Big step step function upgrade for agentic commerce. ucp.dev/2026-04-08/spe… Coming soon to every Shopify storefront.

@fchollet The best models aren’t just correct, they make the signal obvious.

We should view the history of physics as a long-running program synthesis task. Kepler and Newton were searching the space of possible symbolic models to find the simplest one that would best satisfy available observations.

@HarryStebbings Access will be everywhere. The real gap comes from who knows what to do with it.

How do we use AI to help not harm income inequality? "One approach is making sure everyone has ownership, through pension funds or sovereign wealth funds investing in AI. If productivity gains are concentrated, we need mechanisms to redistribute that value so everyone benefits. AI could drive massive advances in energy, materials and infrastructure that raise living standards for everyone." @demishassabis What are the most effective ways we can ensure the gains from AI are distributed across society @pmarca @vkhosla @rcbregman @JonathanRoss321

@collision It’s not a lack of systems, it’s too many without a clear path. That’s where things start to break.

Basically the experience of interacting with US healthcare specialists. Your results are on MyChart. We'll message you on Spruce. Just log into MyPatientPortal. It's right there on FollowMyHealth. Just create an account on myEHR.

@awilkinson The signal changes depending on where you look. Most people don’t have a clarity problem, they have a context problem.

Bahahahha 😂 Nathan Latka’s website: 4.8/5 on TrustPilot TrustPilot website: 2/5 with 181 reviews

TelyAI handles the noise. You handle the decisions. That's what real AI assistance looks like. Visit us at: telyai.org

@rauchg That’s what matters, removing complexity so teams can actually use it without thinking too much.

AI Gateway is quite literally a “peace of mind” product: ✅ No downtime ✅ No lock-in ✅ No keys 🆕 No training

@amasad The shift isn’t just more agents, it’s having a workflow clear enough to actually stay in control of everything.

There’s a reason bootstrapped solo businesses are accelerating on Replit… we gave builders entire teams.

@eladgil @NoPriorsPod @LiamFedus @periodiclabs Interesting shift, but the real value comes when this turns into something clear and usable in practice.

Fun conversation on @NoPriorsPod w @LiamFedus on AI Foundation Model for Atoms from @periodiclabs -Foundation models that understand atoms -LLMs as reasoning engines -Timing to displace AI researchers Liam prior was VP Post Training at OpenAI, now CEO Periodic

The most expensive thing in your workday isn't your tools. It's the time you spend figuring out what actually needs your attention. TelyAI solves that. Before it costs you another hour.

@AliAbdaal @karpathy Surfacing disagreement is interesting, it makes things clearer instead of just adding more summaries.

inspired by @karpathy's LLM wiki idea, i've been building "the brain" - an LLM-maintained knowledge base that compiles raw sources into articles that surface where experts disagree rather than just summarising pipeline: raw files (transcripts, courses, research) → extract specific claims → compile into trust-weighted reference articles. my own content = highest trust, sources defined as "mentors" 2nd highest (ie: people i trust who have good takes), "vetted" vs "unvetted" source material - vetted means i've read it and think it's great, unvetted = research from the internet i haven't yet explored in depth LLM does the bookkeeping (@claudeai sonnet 4.6 seems just as good as opus 4.6 on this), i make the judgment calls on disagreements. question i'm trying to answer: how do i actually make this useful? i'm working on a second book, we run a curriculum for our lifestyle business academy, i make youtube videos, we're building apps - can one knowledge base feed all of it? lots of reusable components - i've got 1000+ youtube videos i've made over the years, thousands of @readwise highlights, tonnes of notes from scientific papers when i was writing feel-good productivity (pre-LLM era)... surely there's a way for our new stuff to benefit from all this so far, the disagreement-surfacing is already the most interesting part. "researcher A says do X, person B says the opposite, here's the evidence for both" is way more useful than a clean summary when you're trying to figure out what to actually teach / write about i'll share more as it develops. if anyone's experimenting with something similar i'd love to hear what's working

@andrewchen Makes sense, you see it faster when someone can think clearly and actually execute.

noticing a trend of startups replacing standard resumes/interviews with week-long (or at least 3-day weekend) in-office trials. Makes sense in a world of AI-generated resumes and interview responses Turns out the best signal for whether someone can do a job is watching them actually do the job. took us 100 years of HR to rediscover apprenticeships!!! 😂

@samgrows Better data helps, but it only matters if the output is actually clear and usable.

BREAKING: Claude can now search YouTube for you! We plugged Algrow directly into Claude so it finally has real constantly updated youtube data No more generic slop from a model that's never watched a video

@Johnsjawn Becoming AI-native is less about tools, more about making outputs clear enough to act on.

Tomorrow we’re breaking down how companies actually become AI-native. Using the AI Transformation Model + a real case study from Ramp. Ben Levick (Head of AI @ Ramp) will talk through how they did it from the inside. If you’re figuring this out right now, you should join. info.notion.so/webinar/webina…

@fortelabs Fast, but without clear structure it’s easy to miss what actually matters.

I asked Claude Code to organize 32 messy files using PARA It read every document, asked smart questions, and sorted them in 4 minutes Accuracy: 78% The 22% it missed? That's exactly why you need to understand the method yourself first

@emollick Strong model, but without clear access it’s hard to turn it into something useful day to day.

Seems like a good model from Meta that is still trailing the current series of releases. The most important thing to note is that it is not open weights. That was the main reason that Meta's models were so important. Without that, it is a lot harder to predict the value of Spark

@mattturck Clarity in judgment is still what turns output into real outcomes.

YES YOU ARE STILL RELEVANT AS A VC EVEN THOUGH CLAUDE PROVIDES BETTER FOUNDER ADVICE AND DOES BETTER DEAL ANALYSIS AND AI AUTOMATES EVEN YOUR SILLY LITTLE MEMES, BECAUSE YOU HAVE TASTE, YOU UNDERSTAND? YOU. HAVE. TASTE. NOW GO WIN THAT 1000x ARR DEAL.

@emollick Delegation works until context gets lost. Clarity in communication is what makes handoffs actually hold.

It is weird that you can approach LLMs as reasonable approximations of humans and get good results, but it is even weirder that you can approach agents as reasonable approximations of organizations (higher ability work is expensive so delegation is important, hand-offs have cost)

@fortelabs Knowing AI isn’t the edge. Clarity on how you think and decide is what makes the output actually useful.

Most people think the way to get better at AI is to learn more about AI I think that's backwards The people getting the most value from AI aren't the ones who know the most about it. They're the ones who know the most about themselves Their psychology, their blind spots, how they communicate, what fills them with energy, their hard-won principles and values. The stuff that takes years of living to accumulate and that no model can guess Without that, AI just gives you the average of the internet. Polished, plausible, completely generic. You accept it because you don't have a clear enough internal compass to push back I learned this the hard way. The more I leaned on AI, the more I lost touch with my own point of view. I couldn't remember what I believed because I'd stopped doing the thinking required to know If you can't articulate what you value, how you make decisions, and what your quality bar is, no amount of computing power will fill that gap. That's not a philosophical point. It's a practical one I run into every single day It's also why the first thing we build in The AI Second Brain isn't a prompt. It's clarity about who you are and what you're trying to do Founding cohort: April 15 to May 1