I have been thinking a lot recently about our dependency on tech (mine, Taught by Humans’ and just in general).

This has been fuelled by a few things:

Getting removed as admin from the Taught by Humans LinkedIn (due to them messing up an account merge they made me do to post a job… Long story!) while Tom the only other person with access to post was on holidays.
Claude going down just before they rolled out Sonnet 4, while I was debugging something simple and had to decide whether to use ChatGPT (I hate it for coding as I can’t get the results I need, it’s often filled with errors), wait for Claude to be back online or do it myself. I opted for the latter, but it got me thinking about why this was even a decision.
Going away with my husband's family to somewhere with “bad Internet and spotty service” which actually meant absolutely no signal or Internet. And how much this bothered me (it was the expectation to be able to use emails and Internet search, rather than absolutely nothing which meant I hadn't downloaded a few things).

Now with my usual “nobody died” so it's all fine mentality, these aren't that big of a deal. I'll be honest, the LinkedIn one and how crap the customer support were is the one which really got me thinking. We use LinkedIn a lot for business - aiming to post three times a week. While we don't get lots of likes and comments, we have run some data and we do get a lot of inbound requests mentioning our LinkedIn posts. So it's important to our sales efforts.

But it is so easy for that to be taken away. Even a change in the algorithm changes who sees our posts. A tech mishap is what caused me to be removed as the only admin. If I was a solo run business, I wouldn't even have had the fall back of having our marketing person with content posting access.

Does LinkedIn Really Matter?

Now in the grand scheme of things, even for our business, LinkedIn is relatively unimportant. If we'd lost access to our Google Workspace, it would have been a lot more stressful. Luckily for us Google is pretty reliable (touch wood) and has a much better customer support process. For others this could be Microsoft, Notion, JIRA, Miro, the list goes on - we save so much of our needed business knowledge and day to day stuff in these platforms.

And how often do we download soft copies to our devices or print hard copies (although I think the environment thanks us). Do we have backups? If a customer session is being run in Miro or ChatGPT, and the platform is down, what happens? If JIRA or Asana (ticketing systems for tech teams to assign work) aren't accessible, when was the last time the sprint was downloaded somewhere else? The strategy doc that took the C-suite hours and hours, what happens if Notion is down for the big presentation? Or forever?

This might sound like some hyperbolic Black Mirror inspired dream, but this last week has seen it happen.

But it Could Matter

The biggest AI news of the week was builder.ai entered insolvency.

Who is builder.ai?

The correct question might be who were builder.ai, but anyway. They are an AI-driven, no-code app development company. A real darling unicorn of the AI world. And the VC world (considering they have raised $500million, including being backed by Microsoft).

This week has shocked the tech world with this apparently $200 million Annual Recurring Revenue company having no money. All 1,500 employees were let go - locked out of buildings and platforms. Even worse, companies were locked out of their applications and cannot access the source code (the way they could run the application elsewhere).

It’s come out that the financials were not exactly factual, and the AI capabilities were over inflated - with much of the work reportedly done manually by underpaid developers in India and other countries (despite claims that it was fully AI-driven).

Why Does This Matter?

One of the biggest issues in the AI adoption space is knowing who to trust (in terms of people and tech). This will cause a lot of companies to, rightly so, doubt what they are seeing. Who else is smoke and mirrors charging a lot of money?

Even more, this raises some issues with the current no-code trend. What happens when we pay for software and they don’t actually own the code? Is this something that needs to be clear when signing up?

One thing it highlights is the need to understand both AI (in terms of what it can and can’t realistically do), and the need for businesses to understand the importance of tech skills (and ownership).

Anyway, I’m off to back up some important docs - just in case.

Claude Down, LinkedIn Locked Me Out - A Reminder About Tech Dependence

Related Articles

Should You be Using Moltbot (formerly Clawdbot) as Your AI Agent?

Why I've Turned Memories Off in ChatGPT and Claude

Humans First: Is AI At Work What We Expected?

Enjoyed this blog?