Read.AI

My team was on a microsoft Teams meeting a few days ago, to discuss some, shall we say, sensitive topics with some potential investors / collaborators. Let’s call the other group Acme International. Someone on their side launched “read.ai” as a meeting participant. At the time I thought it was just someone on their side who had their camera off.

The next day, when my team launched up our own internal Teams meeting to discuss all the juicy details of our meeting with Acme International, we had an uninvited participant in our internal meeting, one “read.ai”. The same no-video participant from the day before. Huh? Wait a minute, did Acme International insert some kind of spyware into our microsoft Teams? We all hastily jumped off that meeting, created a new meeting, we got onto that, and yup, there was “read.ai” as a participant again. We eventually figure out what was going on, and blocked “read.ai” from accessing microsoft Teams.

Here is what we learned so far:

1. During our meeting, one of the Acme International members launched an up-and-coming, super helpful and not at all nosy AI tool which watches and listens to your meeting.

2. By admitting read.ai into the meeting we had given it permission to join other meetings. We do not know yet if that was a EULA feature we “agreed to” by admitting read.ai, or a bug. My guess is it is just typical IT corporate chicanery of doing whatever you want then seeing if you get sued, then wrapping the cost of the suit into cost of doing business, which practice was full developed by Google under the watchful and terrible eye of the Obama administration.

3. After our meeting with Acme International several of us on our side received super friendly emails from a “Michael Clarage via read.ai” asking if anyone wanted to see a summary of the meeting. Of course I never authorized nor wrote any such thing. One of us clicked on this link, and had sign up to read.ai to see the meeting summary, so he signed up, but used one of his junk email accounts, not our company email, and was granted full access to a super helpful and not at all creepy dashboard. This means that anyone with that original link can view the full contents of our private meeting where we discussed some sensitive topics.

4. The not-at-all- creepy dashboard provided a scarily concise and accurate summary of a 1-hour meeting, then a full transcript. You could “chat” with the meeting which means ask the AI questions about the content of the meeting. The dashboard also had a graphical section which showed timeline plots over the 1-hour, where each of us had our own color-coded line which showed our “engagement”, and “sentiment” during the course of the meeting, as well as single numbers between 0 and 100 scoring each of our “engagement” and “sentiment” for the entire meeting. Speculating a bit, maybe because managers don’t have time to look at plots, and for the sake of putting together a list of who to fire this month, it’s much easier to see that once again Michael had a high “engagement” score, but his “sentiment” score was, once again, below the company guideline of maintaining a > 50% sentiment on meetings. I am sure that “engagement” and “sentiment” are measured by running facial recognition and tone of voice through a completely objective and not-at-all hackable LLM and hence CEOs can rest easy that downsizing the workforce is based upon trustable SCIENCE.

5. The dashboard also had a jazzy section for linking all this with JIRA, Salesforce, GMail, other business platforms. I did not see a Quickbooks, but maybe that will be an upcoming premium upgrade. Some well-informed speculation says that ingesting business software development (JIRA), and sales process data (Salesforce), and customer communication (GMail) and eventually financials (QB), will allow our benign AI titans to train the models on ALL aspects of running a business. Then they can offer “small business AI”, which will run your entire business. Think of the savings! Everyone knows that employees are the largest cost of running a business. I’ll go out on a limb, and bet “small business AI” is already being rolled out in Africa because that is the preferred continent for trialing new delights that are being cooked up for the continents which have money.

There’s the facts so far, along with some well-informed speculation. I’m still sitting with what it all means. Inside me are fragments, images of frogs in slowly warming water, of Morpheus saying “we hack into the network and broadcast our pirate signal”, of Aleksandr Solzhenitsyn describing the various tortures for those who resist blissful New World Orders, of cranky old men who can’t use a computer, of the headache of getting your medical records from one office to another, of the knee & hip destroying efforts it takes to actually grow your own food, of someone saying to an older, feebler version of me, in an exasperated tone, “it’s just an iris scan, that’s how we do it now, you can’t get into the hospital without it, you’re being silly….”, of the look on faces high school students who actually cannot, of their own free will, stop using the phone every few minutes,