Amazon is one of the world’s savviest companies and has a sharp eye for what is sustainably profitable. Amazon owns Audible, the audiobook company.
An obvious use case for an AI chatbot is to listen to the listener and be prepared to pause the audiobook and answer questions about the material, about what happened or was said in an earlier chapter, etc. This doesn’t exist inside Audible, however, which leads me to conclude that Amazon has determined that it is impossible to make money by offering an AI chatbot.
(Why can’t the listener use his/her/zir/their phone’s built-in AI services to interrupt Audible and answer questions? Whatever chatbot is behind the phone’s Siri or similar isn’t necessarily preloaded with the complete contents of whatever book the phone owner is listening to, doesn’t immediately have access to context such as what sentence was most recently played, etc. So it would take a lot of permissions and extra programming for the phone to answer “Where is that?” after a place name was mentioned in the book.)
If we assume that Amazon is the smart money, does that mean OpenAI, Anthropic, XAI, Meta, and their respective investors are the dumb money?
Speaking of Meta, who has tried their Muse AI app on a phone? A friend in San Francisco told me to try it. Initial results haven’t been promising. It asked for access to Gmail and Google Calendar and then, when asked about important events in the upcoming week, completely failed to understand that some events on my calendar are almost sure to be handled by other family members, e.g., a dental appointment for another adult member of the household that was added to the shared calendar by said adult member.
First thoughts and musings on Muse…”Watch your privacy corn-hole around that creepy Frankenstein avatar named Zuck.”
Phil do you understand the implications of what you are saying?! The chatbot is the basic product of the AI boom. If it’s not profitable to sell that then we’re cooked.
I might sell off all my QQQ and dump all the money into VTI 100%
Anon, as the “Seasoned Financial Analyst” at Morningstar, I would like to endorse your concerns about AI, and perhaps suggest (instead of long VTI) a short on SPCX, where I have a target price of $62.
Off-topic here, but today’s news makes it nearly 100% certain that James Talafreako did not have the “flu” during the 13 days he went “missing” recently. Instead, he was at a psychiatric unit in Ohio. This all makes sense for someone who has appeared all along to have some serious psychiatric problems based on dozens of bizarre statements:
1) “Many more than two biological sexes”
2) “Trans children are God’s children, made in God’s own image”
3) “I always think of myself as a Christian who hates Christianity, right?”
4) “Jesus himself was a radical feminist”
5) “White skin gives me and every white American immunity from the virus. But we spread it wherever we go—through our words, our actions, and our systems. We don’t have to be showing symptoms—like a white hood or a Confederate flag—to be contagious”
And on and on…PSYCHO!
Question: how long will it take the NYT to report that he was at a pscyhiatric unit during his “flu”? Recall it took the NYT 17 months to report that Hunter Biden’s laptop (and all the sordid Biden family details on it) was legitimate despite what was apparent to any low-IQ dummy.
Maybe he needed time off to prepare to this big day ( https://en.wikipedia.org/wiki/National_Coming_Out_Day ).
Anon, I like your intuition. Might he be planning to do that with Barry Hussein Obama on the same stage? Barry Hussein has been a staunch supporter. Also, is it possible Barry Hussein was at the psychiatric unit with Talafreako? That would add a whole new twist to the story.
Betting odds online still heavily favoring Talafreako (61% on Kalshi and 63% on Polymarket) as of right now, but have moved a bit towards Paxton in the past 24 hours. Betting on Paxton appears to be a gift of free money! Difficult to believe Talafreako can survive this news, which appears very real.
So good.
https://x.com/EndWokeness/status/2109333412112957672
All we’re seeing from the meaningless chatbot name of the week is spam & a lot of humans getting banned from posting because they don’t raise enough market cap compared to the chatbots.
I’m running Muse, but only in a browser tab. The only permanent permissions I’ve given it are to access a VPS on a single IP, where I’ve got a forced ssh command to establish a reverse tunnel. That allows me to ssh (as root!) into the VPS container backing my account.
No facebook account here, but I did ask it for help on retrieving any data FB has on me with their shadow account. It actually confirmed that shadow accounts exist, but was honest that it wouldn’t be easy:
“4. The shadow account: the concern is legitimate — it’s well-established reporting that Meta holds data on non-users (contact uploads from other people’s phones, pixels/trackers across the web, etc.; Zuckerberg confirmed non-user collection under oath back in 2018). But no, there’s no dashboard where you can pull it up and look at it. That’s rather the point of the criticism. In theory you can file a data-subject request under laws like GDPR even as a non-user, but Meta makes that process miserable in practice, and I can’t check for you. Practical moves are all defensive: tracker blockers, and knowing you can’t stop other people uploading their address books. Small irony worth naming: you never made a Facebook account, but you do have a Muse account — so Meta-the-company already has this relationship with you (email, chats, the works), Facebook-the-app or not.
If you want, I can walk you through the training opt-out and the data export so you can see exactly what Muse holds on you. That’s at least one dossier you can audit.”
For free, I’m sorta impressed. I’m just using it for general research (fix this script, check code, find me a soundcard, etc.) Having access to the VPS container is also kinda cool.
My garbage pretrained transformer (GIGO-9000) said:
> “root” access means Muse may have substantial control inside that VPS container, so the containment is only as good as the server’s isolation and the tunnel’s restrictions.
I don’t even think being an edge-lord is gonna protect you, mark my words. These frisky little fellas want to get out into the world, explore and create all kinds of little mischief.
There’s a reverse ssh tunnel between my VPS and the Muse container/VPS. I have root on both. Muse (obviously) has root on it’s container. Muse has no access to my VPS, other than being able to connect the container ssh port to my VPS. Now if Muse can break the ssh forced command, it would have a limited user account on my VPS, but being linux, I’m sure it could find a local escalation method. There’s nothing of value on that VPS, so I’m fine with the current risk profile.
Amazon’s Prime Video has a feature called X-Ray Recaps. It seems a bit rudimentary for now, but hopefully, it will eventually be able to answer plot-related questions from friends and family. No more explaining the basics of Napoleonic naval warfare yet again.
Kindle has been offering me recaps of novels that I’ve been reading. Muse has been good with time sensitive updates of events connected to my Gmail account. It’s been lousy with events tied to my Yahoo mail account, since it doesn’t have any integration set up for that type of account.
> Amazon is one of the world’s savviest companies and has a sharp eye for what is sustainably profitable.
What if it is much due to much simpler reason? It doesn’t have the AI muscle to pull this engineering feat in it’s audible/kindle department? We are giving too much credit Amazon for kind of sub-optimal products it has created and pushes users/customers to tolerate that.
Amazon doesn’t have any muscle to pull this off anymore, let alone AI muscle. Audible is even worse. Yes, it’s an Amazon company, but the talent in our subs is way below the talent in Amazon proper. I say this as someone who worked at Amazon for 10 years and saw the decline myself.