Being an Entrepreneur | June 2026
- Chris Herbert
- 13 hours ago
- 10 min read
Video Summary
Here's a video summary of this month's Round Up articles using Google's NotebookLM
FluentCRM 3.0: AI, SMS, and Owning Your Marketing Stack in WordPress
FluentCRM — a self-hosted marketing-automation plugin used by 70,000+ WordPress sites — shipped its largest release to date, and the headline isn't any single feature but a full rebuild on a modern framework (Vue 3) that makes the dashboard and reporting faster and cleaner.
On top of that foundation, version 3.0 adds three things worth an entrepreneur's attention.
AI features: a set of MCP connectors that let the CRM talk to large language models you already use (ChatGPT, Claude, others), plus AI-assisted email drafting and one-click contact summaries that condense a customer's full history into a readable next-step.
SMS marketing: two-way text campaigns and automations alongside email — abandoned-cart texts, shipping confirmations, billing reminders — sent through Twilio
Gutenberg-native email builder with reusable patterns, so emails are built with the same editor as the rest of a WordPress site.
Bear in mind this is the company's own launch post, not an independent review, and the AI and SMS additions are new enough to be unproven in the field; SMS also carries real per-message costs through the underlying providers.
For entrepreneurs, the strategic point sits underneath the feature list: FluentCRM is self-hosted, meaning you own your contact list and customer data outright, with no per-subscriber SaaS fees that climb as you grow. That's the exact "own the channel" posture the AI search shift is making essential — as Google referrals get less reliable, a direct email and SMS list you control is the audience no algorithm can switch off.
The honest caveat is the trade-off: self-hosting means you are responsible for deliverability, security updates, and maintenance. For a Grey-Bruce operator already running on WordPress, though, it's a credible, lower-cost way to build the direct customer relationships that increasingly decide who gets found.
Date: May 13, 2026 | Source: FluentCRM 3.0: The Next Generation of WordPress Automation (FluentCRM)
Frontier AI Watch: Claude Fable 5: A Frontier Model's Three-Day Life
Disclosure: Mi6 uses Anthropic's Claude in producing this Round Up. The summary below is kept strictly to reported facts and attributed accordingly. In the span of one week, a state-of-the-art AI model arrived for the public and then disappeared by government order. The whiplash is the story — and the lesson.
Part 1 of 2 — The Launch: Frontier-Tier AI Goes Mainstream
On June 9, Anthropic released Claude Fable 5, billing it as its most capable publicly available model and the first of a "Mythos-class" tier — a level above its Opus models — opened to the general public.
The pitch was long-horizon capability: the harder and longer the task, the wider its reported lead. Anthropic cited early-access testing in which Stripe completed a codebase migration across a 50-million-line system in a day, work it said would have taken a team over two months.
Because that capability cuts both ways, Anthropic shipped it with conservative safeguards — queries touching cybersecurity, biology/chemistry, or model "distillation" were automatically routed to its prior model, Claude Opus 4.8, a fallback the company said triggered in under 5% of sessions.
A sibling model, Mythos 5 (the same system with those cyber safeguards removed), was restricted to vetted cyberdefenders. At launch, Anthropic said Fable 5 would be free on paid plans through June 22, then move to metered pricing of $10 per million input tokens and $50 per million output. Worth noting then, as now: the glowing early reviews came from customers Anthropic selected, not independent benchmarks.
Date: June 9, 2026 | Source: Claude Fable 5 and Claude Mythos 5 (Anthropic)
Part 2 of 2 — The Reversal: Disabled by Government Directive Three Days Later
Three days after launch, the picture flipped. According to CNBC, on June 12 Anthropic disabled access to both Fable 5 and Mythos 5 to comply with a U.S. government export-control directive that cited "national security authorities."
Anthropic said it received an order Friday afternoon to suspend all access to the models "by any foreign national, whether inside or outside the United States," including its own foreign-national employees, and that it disabled the models for all customers to ensure compliance — while stating its other models were unaffected. Anthropic said the government did not provide specific details about its concern, apologized to customers, and stated the action did not meet the principles of a transparent, fair statutory process it has publicly supported.
CNBC notes the move follows an earlier clash in which the U.S. Department of Defense labeled Anthropic a supply-chain risk; Anthropic has sued the administration, and that litigation is ongoing.
Date: June 12, 2026 | Source: Anthropic disables access to Fable 5 and Mythos 5 to comply with government directive (CNBC) — Ashley Capoot
WHAT THIS MEANS FOR ENTREPRENEURS
For entrepreneurs — and Canadian ones especially — the whiplash is the takeaway. The directive as reported targets "foreign nationals," which includes Canadians, and in practice all customers lost access (to Fable, not Claude earlier models): any Grey-Bruce operator who started building on Fable 5 during its three-day window has already lost it.
That makes this a live case study for a theme running through this whole edition — don't anchor critical workflows to a single tool or platform you don't control, because access can vanish for regulatory, legal, or geopolitical reasons entirely outside your business.
It's the same "own your infrastructure" logic as the FluentCRM story, applied at the AI layer: keep your AI workflows portable across models, avoid hard dependence on any one of them, and treat frontier capabilities as powerful but potentially impermanent.
Anthropic's other models remain available, so most existing Claude-based work continues — but the episode is a sharp reminder that, in 2026, the AI tools small businesses can be affected by a shifting regulatory ground.
Canada's National Artificial Intelligence Strategy: AI for All
Canada's first comprehensive national AI strategy, a five-year plan launched by Prime Minister Mark Carney, is built on six pillars and a single diagnosis: Canada invented modern AI but ranks among the slowest G7 nations to adopt it.
Only 12 percent of Canadian businesses used AI to produce goods or services between mid-2024 and mid-2025, and just 8 percent of SMEs have adopted it — well behind Nordic leaders. The strategy frames this not as resistance but a "translation problem," noting 78 percent of non-adopters simply don't see how AI applies to their work.
It targets 60 percent business adoption by 2034, roughly $200 billion in GDP gains, and up to 250,000 new jobs, anchored in principles of trust, opportunity, and sovereignty.
For rural Grey-Bruce entrepreneurs and small operators, the concrete supports matter most. The strategy commits a $500 million BDC LIFT financing program for AI tools, a free AI Literacy and Adoption Assessment tool to gauge readiness in a low-risk way, $700 million in affordable sovereign compute for SMEs, and expanded regional delivery through Regional Development Agencies.
Public libraries and community organizations are named as literacy partners for rural, remote, and northern regions — a notable signal for towns outside major hubs.
The agriculture pillar, spotlighting precision-farming firms like Saskatoon's Croptimistic, and the SR&ED plus Productivity Super-Deduction tax incentives offer practical entry points.
The takeaway for local businesses: federal money and tools now exist to close the adoption gap — the advantage goes to those who move first.
Date: June 4, 2026 | Source: Canada's National Artificial Intelligence Strategy: AI for All
AI Search Spotlight
The AI Search Reckoning: Four Reads on the Same Shift
The search bar — the most-used interface on the internet and the front door to most small businesses — is being rebuilt around AI answers. Or is it? The story underneath isn't a single trend; it's a fork. One path absorbs the click; another redirects it. Read these four together: Google's framing, a major publisher's response, the user backlash, and the data that complicates all of it.
Part 1 of 4 — Sundar Pichai on Google Zero and the Future of the Web
In his fifth annual post-I/O interview on The Verge's Decoder podcast, Google CEO Sundar Pichai sat with Nilay Patel and, notably, did not dispute the "Google Zero" trend — the scenario where AI answers questions directly and referral traffic to websites approaches nothing.
He cited Condé Nast's posture of planning as though search traffic doesn't exist (see Part 2), and conceded one live Search result Patel showed him was "more opinionated than it should be." His defence — that referral traffic now comes from many sources and that the clicks that remain are higher quality — is contested by independent research showing AI Overviews correlate with click-through declines that roughly doubled over the past year.
This is the architect of the change telling you plainly the open web is no longer reliable top-of-funnel — and his own concession points the way: thin, generic content erodes first; distinctive, genuinely useful content holds.
Date: May 26, 2026 | Source: How Sundar Pichai is rethinking Google for the AI era (Decoder)
Part 2 of 4 — Condé Nast CEO: Plan As If Search Traffic Will Be Zero
If Part 1 is Google's framing, this is what a major publisher actually did about it. Condé Nast CEO Roger Lynch told his teams to budget "Assume there's no search" — after three straight years in which internal forecasts underestimated how fast search traffic was falling.
He doesn't expect literal zero; he expects search to settle at a single-digit share of total traffic. The cause is visible on the results page itself: where searches once returned a few ads and ten blue links, they now lead with an AI overview, rows of commerce links, and sponsored content, pushing organic results to the second page.
The part that matters most for small operators is Lynch's "barbell effect." Across Condé Nast's portfolio, the brands thriving are at the two ends — large and authoritative (Vogue, The New Yorker) or sharply niche with a loyal, paying audience (Pitchfork). The brands caught in the middle, broad but without deep authority or a devoted niche, are the most exposed. Condé Nast's answer has been subscriptions over clicks: digital subscription revenue grew 29% last year even after material price increases, with retention improving rather than declining.
Editor's note: A small irony worth sitting with — at the time of writing, Condé Nast's own website footer still reads "Copyright 2025." and they don't have a subcription option on their website.. tsk, tsi. If the firm reading the search shift most clearly can leave a stale year in its site furniture, it's a useful reminder for the rest of us: getting the big strategic call right doesn't excuse the small executional details. In a web flooded with AI-generated noise, those details — an accurate footer, a maintained site, a current date — are exactly the low-cost trust signals that tell a human (and increasingly an AI) that a source is real and looked-after. Go check your own footer this week.
Date: May 13, 2026 | Source: Condé Nast CEO: Plan As If Search Traffic Will Be Zero (Search Engine Journal)
Part 3 of 4 — Google's AI Search Just Exposed The Whole Sh*tshow
If Parts 1 and 2 are the institutional view — adapt and move on — this analysis video is the user revolt against being given no choice at all.
The host (a computer-science PhD who breaks down AI developments) argues Google's May 2026 AI Mode overhaul isn't a new feature but a different product wearing the same name: the old model let you decide what to trust from a list of links; the new one hands you a single answer where the only options are accept it or leave.
Her central charge isn't that AI search is bad — it's that Google forced it on everyone and removed the off-switch. The proof is where users fled: not to Bing or Yahoo, but to DuckDuckGo's dedicated "no-AI" search mode, which saw a reported ~30% US install surge the week after launch. People didn't run to a better algorithm — they ran to the one engine that lets them opt out.
Her sharpest line cuts at the gap between adoption numbers and genuine preference: "presence is not endorsement; usage is not enthusiasm." A billion AI Mode users doesn't prove people love it — it may just prove it's hard to avoid when it's the forced default. She backs the revolt with hard signals: Google's US market share posted its biggest single-year drop in over a decade; AI Overviews tested accurate only ~91% of the time (with over half of even the correct answers citing sources that didn't support the claim); and a cited 93% of AI Mode searches ending in zero clicks, quietly defunding the very websites the AI summarizes.
Date: June 2, 2026 | Source: Google's AI Search Just Exposed The Whole Sh*tshow
Part 4 of 4 — The "No-Click Future" Might Be a Little Clicky After All
In a LinkedIn post is the deliberate counterweight to everything above. Marketer Adelle Kehoe, working with Sam Sheridan and Similarweb clickstream data, examined what happened after ChatGPT began surfacing more prominent brand links on May 7th.
The result ran opposite to the doom narrative: referral visits from ChatGPT jumped roughly 150% week-over-week, with about 60% landing on brand homepages, alongside reported uplifts in pageviews per visit (+24%) and time on site (+11%). Kehoe is candid that those engagement metrics are imperfect proxies — and worth noting, Similarweb sells an AI Traffic data module, so it has a commercial stake in the question.
Read against the coverage above, it suggests AI search is splitting in two: Google's Overviews may be absorbing clicks while conversational tools like ChatGPT route users to brands.
Date: May 23, 2026 | Source: Adelle Kehoe on LinkedIn: "The 'no-click future' is looking a little bit clicky"
What This Means for Entrepreneurs and Your Business
Put the four together and the takeaway isn't panic or denial — it's precision and positioning. The old way of being found is breaking, but it isn't being replaced by any single thing, so don't bet the business on one prediction.
First, position at an end of the barbell. You will never be "large and authoritative" at Google's scale — and you don't need to be. The other winning end is a specific niche with a loyal, direct audience, which is exactly where rural and small businesses already have the edge. Deep local or domain authority isn't the consolation prize; it's the position.
Second, respect your customers' choices — the lesson Google ignored. The DuckDuckGo revolt shows what happens when you strip away choice and force your preferred experience on people: a meaningful share quietly finds the exit. It also reveals real, growing demand for human, choice-respecting alternatives — and a warning against reading hype as consensus. "Everyone's on AI now" is a story about defaults, not desire.
Then get concrete: track your referral data by source so you know which side of the fork your traffic is actually on; invest in channels you own — your email list, your site, direct relationships — that no algorithm can switch off; and become the distinctive, genuinely useful source that both humans and AI tools surface by name. The businesses that win the next two years won't out-compute Google. They'll know their audience well enough that they don't have to.
The bottom line: your webite is your owned channel. Make sure you have one, make sure you use it, make sure you track the traffic that hits your website. If you do, your website will possibily impact the other bottom line: revenue, profit and cash flow.
