^{CHITTER CHATTER}

Satya Put Co-Pilot Everywhere. Now He Wants Caution.

Satya Nadella warned businesses that relying too heavily on AI models from companies such as OpenAI and Anthropic could give those providers access to highly sensitive corporate information. Wow, such concern from Microsoft, OpenAI’s biggest commercial partner and one of the companies most aggressively encouraging businesses to pour their operations into AI. Nadella believes that “It's hypocritical of them (AI companies) to freely train on the world’s data while restricting others from doing the same to their models.” Why the caution now, Satya? Maybe he isn’t worried about companies becoming too dependent on AI. Maybe he’s worried they’ll become dependent on the wrong AI companies, and Microsoft will be left supplying the plumbing while someone else owns the customer, the knowledge and the profits. We see you Satya Nadella. Maybe you’re a hypocrite too.

Employees Take a Stand Against Power and Greed

A Google DeepMind researcher resigned after Google entered a Pentagon agreement involving classified military AI work. Google had previously promised not to develop AI for weapons or mass surveillance but removed those commitments from its published principles. Around 600 employees reportedly signed a petition opposing the military deal. I’ve always found their motto of “don’t be evil’ ironic, and I think it’s telling that they’ve made that statement less prominent in 2018, moving it from the top of their Employee Code of Conduct to the bottom.

Tech Hype

Nothing Built. But Somehow Valued at $300M

An AI startup called Elorian has reached a US$300 million valuation before launching a product. Its founder has an impressive DeepMind pedigree, Nvidia’s backing and a plan to advance humanity towards ‘visual AGI’. What it doesn’t yet have is the tedious traditional stuff: a finished product, customers using it or evidence that anyone will pay for it. Real products introduce performance, competition and other inconvenient facts. Until then, Elorian can remain the perfect AI company: richly funded, magnificently ambitious and untroubled by reality.

Tech Fails

AWS Accidently Invents Million Dollar Invoices

The news of AWS suffering a glitch in its billing software made the news everywhere last week. According to this WIRED article, “an error with the cloud computing giant’s billing operation caused some customers’ monthly bills to rise from a few cents to billions of dollars.” As is typical of large tech companies, AWS provided no clear explanation other than to look at their ‘dashboard'. It seems like their billing computation subsystem suffered from really bad hiccups. I know the feeling, happens to me all the time. And like AWS, I also fantasise about billing clients millions for no good reason. At least the company made the news cycle this week. Well done, AWS. Free advertising of your quality tech.

Weird Tech

AI Knows How You Feel. Your Boss Still Doesn’t Care.

Companies are using ‘emotion AI’ to analyse workers’ voices, facial expressions, written comments and even physiological signs of stress. It’s particularly popular in call centres, where software can determine whether employees sound sufficiently calm, friendly and empathetic. To be clear, employers will still ignore your workload, order you back to the office and make half your team redundant, but at least sophisticated technology can now confirm that you’re unhappy about it. The ultimate goal is a workplace where every employee feels heard, measured, categorised, and completely ignored.

^SUBTEXT

The Waffle: “At Elorian, we want to build models that will advance us toward visual AGI.” — Andrew Dai, Elorian founder and CEO

Translation:
“We’re building a machine that will eventually understand the entire visual world. For now, it can already identify venture capitalists at 500 metres.”

The gradual wearing away of corporate principles under sustained exposure to money, power and government contracts.

Valuemaxxing: The science of adding ‘maxxing’ whenever something changes in AI. Right now, it means demanding measurable returns from AI after years of measuring success by how many tokens everyone could burn.

^{THE SHALLOW END}

Are People Scared to Talk ‘Crap’ About Sam Altman?

Amazon abandoned the $40M Sam Altman film Artificial just months after it announced a $50B investment in OpenAI. Neon rescued the movie and wants an Oscar run, but the film reportedly won’t take the traditional festival route. Maybe the question is: “Who’s afraid of Sam Altman?”

Dear Dr Comms
I’m broke but have a brilliant AI idea: a digital clone of yourself that attends meetings, disagrees with everyone and damages relationships on your behalf. How do I pitch it to investors? Yours, Pre-revenue genius.

Dear Genius
Investors love bold ideas, especially when reality hasn’t interfered. Let’s ask the experts:

🧹 Cleaner
“Call it Conflict-as-a-Service and promise it removes unnecessary harmony from the workplace.”

🪚 Sawmill Supervisor
“Pitch it as a productivity product: your clone cuts through meetings, grinds down opposition and leaves you with clean hands when everyone gets annoyed.”

🎈 Children’s Entertainer
“Add the words ‘agentic’, ‘autonomous’ and ‘world-changing’ to your pitch deck. Then make balloon animals until someone signs.”

Layoffs continue but no one wants to call it that.

^{BIN THIS…}

Know someone who lives for this kind of nonsense? Forward this email to them and help me spread the dysfunction.

You can also subscribe to my other newsletter, Lead Different, for a serious take on strategic communications in B2B tech.