13 ms·
“Jassy wrote that employees should learn how to use AI tools and experiment and figure out “how to get more done with scrappier teams.” Isn’t this their genera
by pryelluw 1y ago
“Jassy wrote that employees should learn how to use AI tools and experiment and figure out “how to get more done with scrappier teams.”
Isn’t this their general approach since forever?
- AtlasBarfed 1y ago[flagged]
- pryelluw 1y agoI reason they’ve been past the rectum for a long time and now well into the esophagus. Maybe the goal is a human centipede kind of setup? I mean, one human made up of multiple humans has to be the ultimately productivity machine. Plus they only have to pay one salary …
- happytoexplain 1y agoNote that "scrappier" here doesn't just mean fewer staff, but also less experienced and lower-paid staff.
- dale_huevo 1y agoNext they'll take away the door desks.
- nitwit005 1y agoThat's something odd with the recent AI hype. Companies that were already using AI, are making statements like this. Somehow they want to act like they are making a shift, rather than say they were ahead of the trend.
- chneu 1y agoThat's what businesses do. They want to lower payroll, always, and will use every new innovation to do it. The wording changes, the intention doesn't. If they could pay you nothing they would.
- goatlover 1y agoSounds like a flaw in the dominant economic model.
- ethbr1 1y agoI mean, labor is a significant component of prices for goods. But I expect the increasing income stratification of the 10s+ is a harbinger that we're running out of high-paying jobs for the number of people who are qualified for them. And the window is closing for countries to agree to something like a structural tax on AI with benefits going to society to address the ills. Absent that: further stratification, more employee-less businesses, and not a great future
- nitwit005 1y agoThis doesn't appear to relate to what I said.
- rsynnott 1y agoThe markets seem to like it, so if you go "we're going AI-first!" every six months, you'll get a little stock price boost. Actually _doing_ anything, naturally, is entirely optional. Expect this to repeat until the markets choose a new favourite thing (I'm betting on "quantum"; it's getting a lot of press lately and is nicely vague.)
- tart-lemonade 1y ago"We're going all in on AI" is the modern equivalent to adding .com to your company name to juice the stock price. It's utterly meaningless, yet it works shockingly well.
- droopyEyelids 1y agoSmall and scrappy teams work when the team has less than 8 hours of corporate busywork to do a day (Jira, compliance training, triaging 10k alerts from the new scanning software, etc)
- 827a 1y agoI find it extremely strange that a company leader though it would be ok to just say "our financial situation is in a place where we cannot adequately staff our teams". The market clearly thought it was strange as well, given their stock performance today. Really bad look and poor leadership from Jassy. There's a good way to frame adoption of AI, but this is not it.
- usefulcat 1y ago> The market clearly thought it was strange as well, given their stock performance today. For 6/17, the S&P 500 was down 0.84%, QQQ (Nasdaq stocks) was down 0.98% and AMZN was down 0.59%. AMZN slightly outperformed the market today.
- 827a 1y agoWith AMZN commanding 5% of the S&P's entire market capitalization, the market is not some independent entity that AMZN can be compared to; the S&P follows what AMZN does.
- usefulcat 1y agoI might be misunderstanding, but it sounds like you're claiming that the 95% of the S&P that is not Amazon "follows what AMZN does"? If so, I'd like to hear more about exactly how that works because it sounds very unintuitive, to say the least. In any case, my point was that objectively, AMZN suffered less today than many other stocks, including many other large cap tech (QQQ) and non-tech (S&P) stocks. Considering those facts, it seems like a stretch to claim "the market clearly thought it was strange as well".
- 827a 1y agoYou have what I would describe as a Ben Graham '80s-era view of how the stock market works. The stock market is fundamentally a very different beast post-2008. The top ten companies in the S&P 500 make up 36% of the entire index's market capitalization. In 1980, this number was closer to 15%. One cannot draw any conclusions about how an individual stock in the S&P 10 performs relative to the overall market, because of how correlated these companies are and how much their combined weight contributes to the overall market. Every company in the S&P 10 is a tech company, except Berkshire. They trade together, and how they trade impacts the entire S&P 500. When Jassy says something, it impacts Google's stock. When it comes out that OpenAI might have to sue Microsoft, it impacts Amazon's stock. Why this happens only makes sense to wall street's HFT systems which, quite honestly, are likely closer at this point to ASI than OpenAI; albeit totally unintelligible in their motives and reasoning. Amazon did not outperform the market. The market is Amazon. The S&P 10 is not 10 individual companies; its one company.