Ad Hoc AI Policy & Drama After ‘The Blip 2.0.’ ARD #101

It’s been a week of volatile and ad hoc AI policy and drama after Anthropic’s ‘The Blip 2.0’ — what I’ve been calling the US government’s move against the company’s latest Mythos and Fable 5 models. A nod to the original ‘Blip 1.0’ weekend three+ years ago, when OpenAI’s then-nonprofit board fired and rehired Sam Altman. A full week on from the Anthropic bans, there’s little evidence of resolution — and more signs of ‘on the fly’ demands, priorities and expectations from the White House, especially since the departures of ‘AI Czar’ and tech VCs David Sacks and his #2, Sriram Krishnan. It’s both a ‘hands-off’ and ‘hands-on’ approach to AI — with diminishing benefits for the US and its global tech and AI leadership. Three events I’d like to dig into today for the AI Tech Wave — each with my Take first, then my Overall Take.

(Update: There may signs of a ‘thaw’ from Trump himself on view of Anthropic, as per an interview with Axios. So there maybe a reversal of ‘The Blip 2.0’ yet: “Dario Amodei was ‘nice and smart’ at G7 meeting. So there’s that.)

(1) The White House Pendulum Swings in AI-Governance Strategy

MP TAKE: The administration’s evolving strategy — since the departure of its chief AI Czar and his #2, David Sacks and Sriram Krishnan — is showing signs of an ad-hoc, ‘on-the-fly’ set of vibes. These changing dynamics have big implications for US AI policy, far beyond the factors that drove the Anthropic action.

There are other AI priorities in flux — ranging from US immigration policy that impacts the tech industry’s recruitment of the best tech and AI talent from around the world, to the upcoming trade and tech negotiations with China continuing at the White House this Fall. In addition, the White House is currently considering taking stakes in US AI companies — via some sort of sovereign-wealth-fund vehicle, influenced by OpenAI and others — yet another example of this ad-hoc approach to AI policy. Ironically mimicking ‘Communist’ China’s otherwise soundly-criticized strategy of taking ‘Golden Shares’ in its leading tech and AI companies.

Sources, in narrative order: Axios — Trump’s shadow AI policy. Axios — inside the White House’s AI power center. Washington Post — Trump wants the public to have a stake in AI (inspired by OpenAI). For longtime readers: ‘Nvidia on Air Force One’ in AI-RTZ #1085.

(2) The Administration Wants What It idealistically Wants on Next-Gen AI Models

MP TAKE: Again, these reports highlight the need to balance the technical realities of AI fundamentals and research vs the features one would ideally wish for in a more perfect world. They underline the widening gap between those two necessities.

The administration appears to want a near-100% guarantee that next-gen models cannot be jailbroken — and as every technologist will tell you, that’s near-impossible. These are probabilistic, not deterministic, systems, and almost anything in technology hardware and software, can eventually be jailbroken. It’s a bit like wishing for Santa Claus. And it does not bode well versus a top-down leadership in China driven by engineers and technocrats — against the lawyers and politicians here in the US. It’s increasingly apparent on all matters AI and tech.

Sources, in narrative order: Wired — the White House wants Anthropic to block ALL jailbreaks — it may not be possible. The Verge — Anthropic got hit by export rules nobody understands. The Verge — inside the fight over Claude Mythos 5. Asian Review of Books — ‘Breakneck’ by Dan Wang. For longtime readers: ‘A Failure to Communicate’ in ARD #99.

(3) Does Amazon Like OpenAI More Than Anthropic?

MP TAKE: Anecdotal evidence has the Anthropic Friday-evening ban triggered by a conversation between Amazon CEO Andy Jassy and Treasury Secretary Scott Bessent around ‘AI jailbreak’ concerns. These separate sets of stories again highlight the drama-driven tendencies of the US AI scene — where shifting business interests and political communications at the highest levels can shift AI policies with the slightest change in prevailing sentiments. Not to mention incentives.

Recall that Amazon was an early Anthropic investor ($13B+) and also has a roughly $50B OpenAI investment coming — so it’s vested in both. Yet Amazon-MGM also reportedly pulled a near-finished Sam Altman movie, post that OpenAI investment and data-center deal. This isn’t about money; Amazon’s Anthropic stake alone would be worth hundreds of billions. It’s about business interests, incentives and orientations. No way to run an AI railroad, as I discussed earlier this week.

Sources, in narrative order: WSJ — Amazon CEO Andy Jassy’s talks with US officials triggered the crackdown on Anthropic’s models (aka ‘The Blip 2.0’). Variety — Amazon MGM drops the near-finished Sam Altman movie, post its $50B+ OpenAI investment + AI data-center deal. NYTimes — Amazon’s $75M+ spend on the Melania movie and promotion. For longtime readers: ‘Anthropic’s screeching halt by the US’ in AI-RTZ #1117.

MP OVERALL TAKE

It’s disconcerting to read these reports of how the Anthropic shutdown was triggered by a Jassy / Bessent conversation — especially when neither executive at Amazon nor Treasury is deeply versed in the technical realities of these latest models and their cybersecurity issues. Indeed, indications were that Amazon and Anthropic tech folks were in deep discussions around these ‘jailbreak’ concerns. Not to mention the one-sided, anti-Anthropic views of the Department of Defense and its head, Hegseth.

The current White House set of events seems to be inspired by knee-jerk perceptions of individual threats — real or exaggerated — without a steadier, long-term eye to governing US AI-model safety across the industry. Especially when OpenAI and others are all racing toward AI models that can meet and potentially beat Anthropic’s in the not-too-distant future. And it does not bode well versus a top-down leadership in China driven by engineers and technocrats — against the lawyers and politicians here in the US — especially if these geopolitical, AI-space-race concerns are the true drivers of US AI policy, as presented to the world at large.

Gadget AI — Father’s Day Gift Ideas for Sunday

Sources, in narrative order: The Verge — the Verge’s Father’s Day 2026 gift guide.

Since Father’s Day is this Sunday, I thought we’d focus today’s Gadget AI on Father’s Day gift ideas — the Verge’s guide, plus my own five picks. These are things I own personally, like and actually use, from some of the newest to a few I’ve used for a while. In no particular order:

1. Google Fitbit Air — $100. A screenless health tracker that, with its app, does all kinds of cool health measurements. (The Apple Watch does all this and more, but costs more.)

2. Clicks Power Keys — $99. A MagSafe-powered, BlackBerry-style physical keyboard you clip onto almost any phone — for the retro folks who want real keys and fewer typos. Just got mine and like it a lot.

3. Anker Laptop Power Bank — $120. A powerful 25,000+ mAh bank with a built-in cord that can power a couple of phones, a laptop and more.

4. Alpaka Flight Sling — $49. A small, compact flight sling for all your digital essentials on a plane, from one of my favorite bag makers. (I’m a bag nerd.)

5. Kindle Scribe (2025 models) — $430 to $650. The modestly bigger Kindles, now in three new models (two black-and-white, with and without front light, and a color one). Read your Kindle books, write on infinite notebooks with the pen — a delight if you’re in the Kindle ecosystem.

MP Take: A handful of Father’s Day gadget ideas for Sunday — across health, productivity, power, carry and reading. They all have AI tinges, since they power or enable devices that are increasingly AI-capable — thus Gadget AI. Things I own and actually use.

Questions

Q1 — What is MP’s favorite and most-used of the five?

The Anker power bank. A must-have in today’s world of powered gadgets — for AI and otherwise. And there’s a regulatory wrinkle: airlines are increasingly clamping down on how many power banks you can carry, with a number now down to just one or two. If you’re down to one, the Anker I recommended packs enough capacity (20,000+ mAh) to run two or three devices, with a built-in USB-C cord and a relatively compact body. That’s my one pick to travel with.

Q2 — Which pick has the biggest wished-for improvement?

The Google Fitbit Air. As someone who uses the Apple Watch for health and everything else, I like the Fitbit Air — but I’d love for it to have a small clock or watch display. It’s one reason I also use a similarly priced health band from Xiaomi, which has a beautiful small OLED display. If a gizmo is going to take up your wrist real estate, the natural habit is to glance there for the time. So that’s my wish-for feature.

Full Source Reading —

For the broader context, see the canonical sources for ARD 101 — in today’s narrative order:

Event 1 — The White House Pendulum Swings in AI-Governance Strategy

• Axios — Trump’s shadow AI policy

• Axios — Inside the White House’s AI power center

• Washington Post — Trump wants the public to have a stake in AI (inspired by OpenAI)

• AI-RTZ #1085 — Nvidia on Air Force One

Event 2 — The Administration Wants What It Wants on Next-Gen AI Models

• Wired — The White House wants Anthropic to block ALL jailbreaks (may not be possible)

• The Verge — Anthropic got hit by export rules nobody understands

• The Verge — Inside the fight over Claude Mythos 5

• Asian Review of Books — ‘Breakneck’ by Dan Wang

• ARD #99 — A Failure to Communicate

Event 3 — Does Amazon Like OpenAI More Than Anthropic?

• WSJ — Amazon CEO Andy Jassy’s talks with US officials triggered the crackdown on Anthropic’s models

• Variety — Amazon MGM drops the near-finished Sam Altman movie

• NYTimes — Amazon’s $75M+ spend on the Melania movie and promotion

• AI-RTZ #1117 — Anthropic’s screeching halt by the US

Gadget AI — Father’s Day Gift Ideas

• The Verge — The Verge’s Father’s Day 2026 gift guide

• Google Fitbit Air

• Clicks Power Keys

• Anker Laptop Power Bank

• Alpaka Flight Sling

• Kindle Scribe (2025 models)

Shorts Clips from today

Clip 1 — AI Jailbreaking: Impossible to Guarantee

Watch on YouTube Shorts

Reports say the White House wants a near-100% guarantee that next-gen AI models cannot be jailbroken. Every technologist will tell you that’s near-impossible — these are probabilistic, not deterministic, systems, and almost anything can be jailbroken.

MP Take: There’s a fundamental disconnect between the policy wish for zero jailbreaks and the technical reality — it’s a bit like wishing for Santa Claus. That widening gap between what’s asked for and what can be delivered does not bode well against China’s engineer-led, top-down approach.

Clip 2 — Top Travel Power Bank: Anker

Watch on YouTube Shorts

MP’s favorite and most-used of his five Father’s Day picks: the Anker laptop power bank (about $120, 20,000+ mAh, built-in cord). It can power a couple of phones, a laptop, an Apple Watch and a Kindle.

MP Take: A must-have in today’s world of powered gadgets, for AI and otherwise. And airlines are clamping down — many now limit you to one or two power banks — so a single high-capacity unit that runs two or three devices is exactly what you want for travel.

Clip 3 — Wish for Fitbit Air: A Watch Face

Watch on YouTube Shorts

MP’s biggest wished-for improvement among his five Father’s Day picks: the $100 screenless Google Fitbit Air. He’d love it to add a small display for the current time.

MP Take: If a gizmo is taking your wrist real estate, the natural habit is to glance there for the time. It’s why I also use a similarly priced Xiaomi health band with a small OLED display. Give the Fitbit Air a little clock and it’s even better.

Clip 4 — Amazon’s OpenAI vs Anthropic Dilemma

Watch on YouTube Shorts

Amazon is an early Anthropic investor ($13B+) and also has a ~$50B OpenAI investment coming. Anthropic’s ‘Blip 2.0’ ban was reportedly triggered by a Jassy / Bessent conversation, and Amazon-MGM also dropped a near-finished Sam Altman movie.

MP Take: This isn’t about money — Amazon’s Anthropic stake alone is worth hundreds of billions. It’s about business interests, incentives and orientations. The anecdotal evidence is building, and these motives will get sorted out for the history books.

About AI Ramblings Daily (ARD), and AI-RTZ

Both are daily. Both are free. Both are about AI. But they’re different mediums carrying different messages.

AI-RTZ is the morning text — a deeper written take on one idea, published by at least 5 AM EST. Today: post #1122.

AI Ramblings Daily is the afternoon video + podcast — my ad hoc takes and perspective on the day’s AI issues & news flow, around 20 minutes, with short 1-2 minute clips for quick topic views. Today: episode #101.

Subscribe to either or both on michaelparekh.substack.com. They run as separate Sections you can opt into or out of.

Links used in today’s show (already embedded inline above; listed here for reference)

Take 1 — The White House Pendulum Swings:

• Axios — Trump’s shadow AI policy

• Axios — Inside the White House’s AI power center

• Washington Post — Trump wants the public to have a stake in AI (inspired by OpenAI)

• AI-RTZ #1085 — Nvidia on Air Force One

Take 2 — The Administration Wants What It Wants on Next-Gen AI Models:

• Wired — The White House wants Anthropic to block ALL jailbreaks (may not be possible)

• The Verge — Anthropic got hit by export rules nobody understands

• The Verge — Inside the fight over Claude Mythos 5

• Asian Review of Books — ‘Breakneck’ by Dan Wang

• ARD #99 — A Failure to Communicate

Take 3 — Does Amazon Like OpenAI More Than Anthropic?:

• WSJ — Amazon CEO Andy Jassy’s talks with US officials triggered the crackdown on Anthropic’s models

• Variety — Amazon MGM drops the near-finished Sam Altman movie

• NYTimes — Amazon’s $75M+ spend on the Melania movie and promotion

• AI-RTZ #1117 — Anthropic’s screeching halt by the US

Gadget AI — Father’s Day Gift Ideas:

• The Verge — The Verge’s Father’s Day 2026 gift guide

• Google Fitbit Air

• Clicks Power Keys

• Anker Laptop Power Bank

• Alpaka Flight Sling

• Kindle Scribe (2025 models)

Q1 + Q2 — MP’s favorite of the five + biggest wished-for improvement:

• (no external sources — MP’s own analyst view)

Companion text:

• AI-RTZ #1122 — OpenAI and Anthropic accelerate their enterprise salesforces

AI Ramblings Daily on AI-RTZ is here to think through AI and reset. Together.

Today’s AI-RTZ #1122 — OpenAI and Anthropic accelerate their enterprise salesforces — OpenAI and Anthropic are both beefing up their professional sales forces against Microsoft, Salesforce and everyone else, accelerating their path to tens of billions in revenues ahead of their mega-AI IPOs. Recommended as today’s reading post.

Look out for this weekend’s Saturday Weekly Roundup, and Sunday’s ‘Bigger Picture.’ Have a great weekend — and a Happy Father’s Day to all.

Thanks for joining us today, AI Curious Folk. Stay tuned.

(NOTE: The discussions here are for information purposes only, and not meant as investment advice at any time. Thanks for joining us here.)

Subscribe to AI: Reset to Zero for daily AI Ramblings + Sunday Bigger Picture posts.