Peter Dao

Humanity, the organic bootloader for Artificial Intelligence.

In the field of artificial intelligence, we are currently at the deep blue moment. We’re essentially brute forcing everything with large language models. Scaling it to a point where it eclipses all of our own inputs and multiples of those in inputs. But I don’t think that gets to the underlying problem about the nature of reality. But if it solves 99% of the problems and is the gateway to the next step, then let’s just do it.

Every single one of these hires have been a senior role. I don’t see any entry-level mid-level employees running away from Google. There is so much talent in the pipeline that I am certain the OG’s are fearful. That’s why they need to capitalize now. There’s a Dario, Demis, etc type doing BS work as we speak.

@ashwingop You can easily convey this idea a lot better in your own voice. Writing just isn’t one of the strengths of a LLMs at the moment. Ironically, because they’ve absorbed all of English literature, yet still lacks significant taste. Write it yourself.

@DaoPeter Care to add something constructive?

Claude Tag is a Trojan horse.  Not because Anthropic is doing anything evil. Because the incentives are obvious. Day one, this looks like a great feature: tag Claude in Slack, let it follow the thread, remember context, connect to tools, break down tasks, chase work, and act like a teammate. But that is exactly the problem. The moment your AI vendor becomes a shared coworker, it stops being just a model provider. It starts becoming the place where work is interpreted, remembered, routed, and eventually executed. That is not model lock-in. That is context lock-in. You are now renting your company back from them. Models can be swapped. Agents can be copied. But the memory of how your company actually works is much harder, maybe impossible, to move: the Slack scar tissue, the exception paths, the customer promises, the unfinished threads, the weird workflows, the implicit owners, the “we tried that in Q2 and it failed” knowledge. Once that lives inside one vendor’s agent layer, you are not renting intelligence anymore. You are renting your company’s operating memory. And the pricing model makes it even more dangerous. A human coworker has a salary. Claude has unbounded tokenized activity. The more work moves through it, the more the vendor captures not just IT spend, but labor spend. This is the enterprise bargain people will regret: Convenience now, and rapid decent into dependency. The right architecture is simple: rent the best intelligence from whoever is best this month. OpenAI, Anthropic, Gemini, open source, whatever. But own the context layer. Your company memory should be inspectable, permissioned, portable, and model-neutral. It should not be buried inside the same vendor that sells you the intelligence and the workflow surface. Claude Tag is useful. That is why it is dangerous. Rent the intelligence, but own the context. Or, regret later.

@SleeperRangers ERA is overblown, I need three good pitches that you can alternate and lean on. I need to see you attacking hitters and throwing a good mix. If you can do that, you can make it in the show.

The Rangers are calling up their No. 5 prospect, Winston Santos. He’ll be seeking his major league debut. Santos has made 11 starts in Double-A and has a 7.44 ERA with 58 strikeouts. Jose Corniell has been optioned to Triple-A.

@AndrewCurran_ @giffmana You simply just had to wear a hat to a UFC fight and enjoy some crazy KOs, along with making a few donations. Should’ve taken no more than 72 hours to resolve this.

Soon.

it still took about 20 to 30 iterations with the latest model to get something that has a decent amount of taste and functionality. Regressing to an older model just made that about 3 to 5 times harder.

Can anyone catch up to a model that was released in February in the middle of 2026? They said open source was at least a year behind but I don’t even see local competition scratching what was just unleashed. Two weeks in artificial intelligence is nearly a few months.

@karpathy @alexalbert__ Where's Fable?

This is a new paradigm for interacting with Claude that is significantly more "inline" with all the other human activity org-wide. Once you do all of the under the hood engineering work to make this "just work" (e.g. across tools, integrations, compute environments, memory, security, etc.), Claude basically joins the team in a seamless way - you can talk to it as you would talk to a person and it can help with a very large variety of workloads. Imo this is the 3rd major redesign of LLM UIUX. The first paradigm was that the LLM is a website you go to, the second was that it is an app you download to your computer. This third one is that it is a self-contained, persistent, asynchronous entity with org-wide tools and context, working alongside teams of humans. It really takes a while to wrap your head around it, but it works and it is awesome.

@TexasMike2014 They don't actually play poker lol. no real player gets mad about this, its negative EV for you and basically free money for anyone willing to gamble. Would be an honor to have you at my table even for a blind all-in.

Man people are really going crazy about the dark all ins. There’s 1 tourney with unlimited reentries, the rest are almost all 2 bullets/freezeouts. I can’t play as many hours as everyone with my heart failure, I’m able to rest for extra hours and still use max bullets. Not sorry

people will say anything to trade off the arbitrage on a prediction market , what a time to be alive!

Rangers lineup for June 23, 2026 at Miami

@bcherny Can you please work on the transcription on all models? Frankly, I’m tired of typing and the newest model can really just get it. I just need a reliable way to transcribe from voice to text. I currently use a third-party transcription service and have to port it manually over.

Claude is really proactive with Claude Tag. You don’t need to prompt it to do work, it can do work proactively based on your instructions. It has excellent memory and access to your data, so it can behave differently per channel

Whoever releases Fable-level first gets my vote. Pretty simple really. The qualitative side.

🚨 SCOOP(s): - GPT-5.6 has been delayed and will no longer release this week. New target is ~mid-July. - DeepMind are not satisfied with the current state of 3.5 Pro and it will no longer launch this month. - Preparations for the launch of Bidi, OpenAI's new voice model, are underway in ChatGPT and we could see it available as soon as this week. - Claude Sonnet 5 is currently available for select enterprise customers under an Early Access Program and is seen as a stop-gap as progress on getting Mythos/Fable 5 back out have stalled. A bit of a disappointing end to the month, but July should prove more fruitful!

@kennlandry very bad AB. You have to hold til he paints one

The Rangers have the bases loaded with two outs for Josh Jung

@Rangers The division is ours! We just need to capitalize versus under performing teams and we will have ourselves a playoff birth

Great start to the road trip! #AllForTX

Good eye Wyatt, he's seeing it well right now

Wyatt comes thru here, he's fully back.

Gotta give credit to Skip for these bullpen starts to get Rocker fired up. He pitches with more heart when he's forced to start off the bench, love it.