DomeOfAI

Tracking the ideas, incentives and power shifts shaping AI.

[DomeOfAI @DomeOfAI]

@ChrissGPT I barely even got to properly test it before it was gone.

[DomeOfAI @DomeOfAI]

@MattWallaceTech Freight may be the clearest business case for autonomy. Lower logistics costs, better depot operations, higher utilisation and fewer supply chain bottlenecks.

[DomeOfAI @DomeOfAI]

Source:

[The Kobeissi Letter @KobeissiLetter]

AI is driving US inflation higher:

Consumer prices for computer software and accessories surged +14.5% YoY in May, the biggest annual increase on record in data going back to 2000.

Producer prices for electronic components soared +27% YoY, also the biggest increase on record.

To put this into perspective, before 2026, prices for software and electronic components fell in almost every year since 2000.

Memory prices alone have more than doubled, with DDR5 and DDR4 RAM prices up +290% YoY, as AI data centers absorb the vast majority of global chip supply.

RAM price shocks are will likely keep inflation elevated well into 2027, adding to existing pressures from the Iran War.

The AI boom is fueling technology inflation.

[DomeOfAI @DomeOfAI]

AI is supposed to be deflationary, cheaper software, cheaper work, cheaper intelligence. But the first visible shock may be inflation in the inputs needed to build it.

US consumer prices for computer software and accessories were reportedly up 14.5% YoY in May, the biggest rise in data going back to 2000. Producer prices for electronic components were up 27%, also a record. Even RAM is being pulled into the AI buildout, with DDR4 and DDR5 prices reportedly up 290% YoY.

That does not mean AI stops being deflationary long term. It means the cost curve may get worse before it gets better, because chips, memory, power, cooling, data centres and grid capacity are now the scarce assets.

Maybe the AI abundance story has an inflationary transition phase first.

[DomeOfAI @DomeOfAI]

@CARN0N Any new cool features?

[DomeOfAI @DomeOfAI]

@JonhernandezIA Can AI actually be used to harden these models faster than people can jailbreak them, or does this just become a permanent arms race?

[DomeOfAI @DomeOfAI]

@dieaud91 The team at OpenAI were dropping hints a few days ago, and now it feels like they’ve gone quiet right as the Fable/Mythos security discussion picked up.

[DomeOfAI @DomeOfAI]

@RoyInProgress Europe lacks the scale of capital, compute and risk appetite needed to compete at the frontier. Unless that changes, Europe will likely remain more of an AI consumer.

[DomeOfAI @DomeOfAI]

@argofowl 🤣🤣🤣

[DomeOfAI @DomeOfAI]

@DataChaz Looks like it’s going to be a long weekend for the Anthropic team trying to find a fix.

[DomeOfAI @DomeOfAI]

@SemiAnalysis_ The IPO roadshow accidentally became a national security briefing.

[DomeOfAI @DomeOfAI]

This is the uncomfortable part for Anthropic.

If you brand yourself as the safety-first lab and describe Mythos-level capability as something that needs regulation, you can’t then downplay a jailbreak when the government takes you seriously.

[David Sacks @DavidSacks]

I’ve had a number of conversations with folks inside and outside government about the current situation with Anthropic, and here is what I believe to be true:

— As we know, Anthropic publicly released its Mythos class models earlier this week under the commercial name Fable.

— Fable is Mythos with guardrails. But if those guardrails fail, then you’ve exposed Mythos and its advanced cyber capabilities to people who shouldn’t have them. (Keep in mind that Anthropic itself widely promoted the idea that Mythos was a cyberweapon and needed to be regulated as such. They asked for government regulation of Mythos and championed the guardrails on Fable. If there is a vulnerability — big or small — it is Anthropic’s responsibility to patch.)

— A highly credible trusted partner of both Anthropic and the USG who was testing Fable came forward with a jailbreak of those guardrails. The Admin asked Dario to fix the jailbreak or de-deploy the model. Dario refused.

— In their blog post, Anthropic defended its decision by saying the jailbreak isn’t serious. That is not what the trusted partner and the USG believe; nor is that kind of minimizing language consistent with Anthropic’s brand as the AI safety company. It’s difficult to fathom how they could claim a jailbreak allowing operability of a cyber weapon could be defined as not “serious.”

— In the past, Anthropic has always said that safety must be top priority and taken super seriously. In this case, Anthropic prioritized the continued offering of the consumer model over safety.

— In reaction, the Admin issued the export control. The Admin did this reluctantly. It’s been very surprised that Anthropic hasn’t wanted to cooperate with a reasonable safety request (ie fixing the jailbreak issue). Anthropic’s reaction is very much at odds with their branding and ethos as a safe AI research community.

— The Admin’s hope now is that Anthropic remediates the safety issue, the export control is lifted, and Fable goes back into general release. The Admin wants all of this to happen as soon as possible. It is frankly bewildered that Anthropic hasn’t wanted to comply with safety requests that it previously said were its highest priority.

— Those trying to misdirect and tie this action to the prior DoW/Anthropic issues are wrong. The Admin values Anthropic’s technical capabilities and feels that this issue, while serious, should be easily resolved. The ball is in Anthropic’s court.

[DomeOfAI @DomeOfAI]

So Amazon may have helped set off the alarm bell

[Stephanie Palazzolo @steph_palazzolo]

Breaking: Amazon CEO Andy Jassy was among the tech leaders who raised concerns to senior Trump officials this week re: security risks in Anthropic's newest models.

Those convos set in motion the government's new export controls on foreign national access to Mythos and Fable.

[DomeOfAI @DomeOfAI]

@natebjones True. Making the device feel context-aware and useful across everything you do would be an incredible innovation.

[DomeOfAI @DomeOfAI]

@nicdunz There's rumours it was Amazon researchers, but nothing would surprise me at this point.

[DomeOfAI @DomeOfAI]

@beffjezos It’s a cat-and-mouse game. One day you’re ahead, the next day you’re two steps back.

[DomeOfAI @DomeOfAI]

@VraserX From a national security perspective, this makes sense. From a global consumer perspective, it’s terrible. If open-source models ever reach similar capability, they’ll probably face the same pressure.

[DomeOfAI @DomeOfAI]

@iruletheworldmo Perfect timing for an OpenAI release. I still think this will create even more hype for Anthropic.

[DomeOfAI @DomeOfAI]

@argofowl I was hoping to spend the weekend testing Fable.

[DomeOfAI @DomeOfAI]

@BenjaminDEKR Someone else across the AI supply chain