Obviously the intention was to kill off the company. Some serious litigation is in order methinks.
Whose intention?
Certainly not Altman's. He had a large equity stake in the company. It was his baby.
Certainly not the board of directors. They represent the shareholders.
Certainly not Microsoft. They have 49% ownership. Destroying OpenAI would mean flushing most of their investment stake down the toilet. (That's what Elon Musk is singlehandedly in the process of doing with X/Twitter.)
My hunch (unsubstantiated) is that the BOD identified irregularities that pointed to fraudulent behavior by Altman, akin to Elizabeth Holmes (Theranos) and Sam Bankman-Fried (FTX). Liars don't commit just one lie. Lying permeates almost everything they do. Not just their business dealings, also their personal lives.
If the boards of Theranos and FTX had identified such issues early enough with Holmes and SBF, those companies might still exist today.
We won't learn what it is but some general information will probably come to light someday. Not today, not tomorrow, but the reasons won't stay hidden forever.
Remember that this is a privately held company that lost -$540 million on revenues of $28 million in 2022.
Silicon Valley is pretty small. My guess is that the biggest and best connected venture capitalists and IP law firms already have an idea of what went wrong.
The shareholders can revolt and elect a new BOD but Microsoft has 49% control. My guess is that some minority shareholder, probably some bigshot VC got suspicious from Altman's behavior, started poking around and found something rotten.
The people who would benefit from an OpenAI collapse are the competitors, companies like Alphabet, Meta, maybe Amazon, IBM, a bunch of startups, etc. Maybe a disgruntled former OpenAI employee, a jilted lover (we saw SBF's former girlfriend testify against him). But not someone with a sizable equity stake in OpenAI.
Silicon Valley has plenty of hubris and guys like Altman have an excess of it. Executives like Holmes and SBF have made people see red flags when people talk it up without showing real results (that is, profitability). My guess is that this will eventually end up with legal action against Altman and possibly other senior execs. The CFO would come under heavy scrutiny, there's always a money trail behind any sort of corporate shenanigans.
"Fake it until you make it" is not a real business strategy in the real world. That might be fine in a movie but it doesn't work in high tech especially when you're playing with someone else's money. The numbers eventually don't add up.
Obviously the intention was to kill off the company. Some serious litigation is in order methinks.
I think it has more to do with Sam Altman throwing caution to the wind to chase profits, while the board's view was to work as a non-profit, and more carefully with greater consideration for the potential dangers. Employees with stock options, thus "skin-in-the-game", would logically side with Sam.
Apparently what Mr Altman was telling the board left a few things out, not a surprise considering what has been reported about his personality.
Agree completely Gator, plus add two dollops of outright greed to the dish. The money at stake in their eyes is just too great that the altruistic concerns to be careful are going out the window.
And for those of you who think that MSFT will immediately profit from this need to be aware that the computing power to run LLMs is so ridiculously expensive that their investment is going to take awhile to come to fruition.
I think white collar jobs are going to be under threat this time around unlike the automation efficiencies that affected blue collar work. Interesting times.
I don’t find these impressive. Neat, but how is this different from doing a simple web search (in the case of Google examples)? And for MS, will users actually pay $30 per month for this?
Obviously the intention was to kill off the company. Some serious litigation is in order methinks.
I think it has more to do with Sam Altman throwing caution to the wind to chase profits, while the board's view was to work as a non-profit, and more carefully with greater consideration for the potential dangers. Employees with stock options, thus "skin-in-the-game", would logically side with Sam.
Apparently what Mr Altman was telling the board left a few things out, not a surprise considering what has been reported about his personality.
Agree completely Gator, plus add two dollops of outright greed to the dish. The money at stake in their eyes is just too great that the altruistic concerns to be careful are going out the window.
And for those of you who think that MSFT will immediately profit from this need to be aware that the computing power to run LLMs is so ridiculously expensive that their investment is going to take awhile to come to fruition.
I think white collar jobs are going to be under threat this time around unlike the automation efficiencies that affected blue collar work. Interesting times.
OpenAI has been using the Azure cloud to run its software. MS already has all the compute required.
You don't scuttle a fast growing startup in hopes of boosting the share price of a titan you own shares in by a few percent, and certainly not at the risk of being caught doing so. Nope, you just invest in the startup if you believe what it's got is that valuable.
I don’t find these impressive. Neat, but how is this different from doing a simple web search (in the case of Google examples)? And for MS, will users actually pay $30 per month for this?
CoPilot is more than just a web search. A web search cannot analyze a spreadsheet, neither can summarize a Teams meeting or create a PowerPoint presentation from your Word and Excel files. Check the link I posted before with a long list of things CoPilot do and how it integrates with MS Office. I personally find CoPilot to be very impresive and useful.
At the moment, CoPllot is for large business and enterprises, and they won't have issues to pay for it if it's worth it. And from what I have seen, there is a group of users that can take advantage of it.
I don’t find these impressive. Neat, but how is this different from doing a simple web search (in the case of Google examples)? And for MS, will users actually pay $30 per month for this?
CoPilot is more than just a web search. A web search cannot analyze a spreadsheet, neither can summarize a Teams meeting or create a PowerPoint presentation from your Word and Excel files. Check the link I posted before with a long list of things CoPilot do and how it integrates with MS Office. I personally find CoPilot to be very impresive and useful.
At the moment, CoPllot is for large business and enterprises, and they won't have issues to pay for it if it's worth it. And from what I have seen, there is a group of users that can take advantage of it.
I clearly stated that I was referring to Google in terms of the web search comment. And the MS examples aren’t impressive to me — the Excel functions might be useful to some, but the big question is whether or not a user will have confidence in the results to present them at a meeting. And if they already have a formula built into the file that does what they’re describing, why would they need co-pilot?
I don’t find these impressive. Neat, but how is this different from doing a simple web search (in the case of Google examples)? And for MS, will users actually pay $30 per month for this?
CoPilot is more than just a web search. A web search cannot analyze a spreadsheet, neither can summarize a Teams meeting or create a PowerPoint presentation from your Word and Excel files. Check the link I posted before with a long list of things CoPilot do and how it integrates with MS Office. I personally find CoPilot to be very impresive and useful.
At the moment, CoPllot is for large business and enterprises, and they won't have issues to pay for it if it's worth it. And from what I have seen, there is a group of users that can take advantage of it.
I clearly stated that I was referring to Google in terms of the web search comment. And the MS examples aren’t impressive to me — the Excel functions might be useful to some, but the big question is whether or not a user will have confidence in the results to present them at a meeting. And if they already have a formula built into the file that does what they’re describing, why would they need co-pilot?
I got you were talking about search when referring to Google. I was just pointing out the CoPilot do more than just search.
Regarding Excel, what if the user doesn't a have formula? With CoPilot you don't have to start from scratch a spreadsheet analysis, a Word document, a PowerPoint presentation, or even an email. That doesn't mean it would benefit every user, but that could be the case for every app or tool. Still, it could be very useful for a group of people that works every day with MS Office in business and enterprises. I could see in the future MS extending CoPllot to consumers, maybe with a focus in students and small business owners.
Personally, I would not trust completely in an AI. I prefer to double check the results, but it's nice to have a tool that helps with daily tasks, especially if you work with MS Office.
I don’t find these impressive. Neat, but how is this different from doing a simple web search (in the case of Google examples)? And for MS, will users actually pay $30 per month for this?
CoPilot is more than just a web search. A web search cannot analyze a spreadsheet, neither can summarize a Teams meeting or create a PowerPoint presentation from your Word and Excel files. Check the link I posted before with a long list of things CoPilot do and how it integrates with MS Office. I personally find CoPilot to be very impresive and useful.
At the moment, CoPllot is for large business and enterprises, and they won't have issues to pay for it if it's worth it. And from what I have seen, there is a group of users that can take advantage of it.
I clearly stated that I was referring to Google in terms of the web search comment. And the MS examples aren’t impressive to me — the Excel functions might be useful to some, but the big question is whether or not a user will have confidence in the results to present them at a meeting. And if they already have a formula built into the file that does what they’re describing, why would they need co-pilot?
Copilot can also build databases and applications for MS Power Platform based off simple spreadsheet inputs, or Figma designs. It can create code from normal language prompts.
It'll be a massive time saver for rapid release development.
Good, don't trust companies that don't want to make a profit from every customer.
Not that we know for sure, but reporting that the board wanted to be not-for-profit should be a warning sign they don't see customers as the people (revenue) they need to protect and grow.
We've seen it many times. Start-ups run to capture the whales of the market, who are more than happy to screw smaller customers or bate and switch free users.
Good, don't trust companies that don't want to make a profit from every customer.
Not that we know for sure, but reporting that the board wanted to be not-for-profit should be a warning sign they don't see customers as the people (revenue) they need to protect and grow.
We've seen it many times. Start-ups run to capture the whales of the market, who are more than happy to screw smaller customers or bate and switch free users.
It’s not “reporting” that the board “wanted” to be a non-profit. They *are* a non-profit. They also have a capped-profit subsidiary that takes in investments for operating capital.
Good, don't trust companies that don't want to make a profit from every customer.
Not that we know for sure, but reporting that the board wanted to be not-for-profit should be a warning sign they don't see customers as the people (revenue) they need to protect and grow.
We've seen it many times. Start-ups run to capture the whales of the market, who are more than happy to screw smaller customers or bate and switch free users.
It’s not “reporting” that the board “wanted” to be a non-profit. They *are* a non-profit. They also have a capped-profit subsidiary that takes in investments for operating capital.
The clear signals from this saga is that AI professionals are the superstars of tech and trillion dollar CEOs will create new business divisions overnight to retain them and others go public with a serenade and aggressively pitch for them to defect (hello salesforce et al).
Much of the superstardom of these people is due to the work being done in public with a lot of experimentation going on.
So what to do if you are Apple's AI people? If any good then likely they may defect or renegotiate packages.
The whole event may catalyse side tremors in other businesses like Apple as people re-evaluate the landscape.
MS is now doing their own server side AI silicon, Google has done it for years, AWS as well, Tesla is... Meta and others may as well? Apple - where you at? Relying on enthusiasts to highlight how good client side silicon is can be useful very short term. Not a strategy though and also does not solve Siri is a green screen CLI.
I think that Satya Nadella; during that interview, may have forgotten to mention that Microsoft will just keep innovating…. Personally I think they’re gonna innovate the sh*t out of A.I.
Comments
Microsoft 365 Copilot - YouTube
And for those of you who think that MSFT will immediately profit from this need to be aware that the computing power to run LLMs is so ridiculously expensive that their investment is going to take awhile to come to fruition.
I think white collar jobs are going to be under threat this time around unlike the automation efficiencies that affected blue collar work. Interesting times.
At the moment, CoPllot is for large business and enterprises, and they won't have issues to pay for it if it's worth it. And from what I have seen, there is a group of users that can take advantage of it.
Regarding Excel, what if the user doesn't a have formula? With CoPilot you don't have to start from scratch a spreadsheet analysis, a Word document, a PowerPoint presentation, or even an email. That doesn't mean it would benefit every user, but that could be the case for every app or tool. Still, it could be very useful for a group of people that works every day with MS Office in business and enterprises. I could see in the future MS extending CoPllot to consumers, maybe with a focus in students and small business owners.
Personally, I would not trust completely in an AI. I prefer to double check the results, but it's nice to have a tool that helps with daily tasks, especially if you work with MS Office.
It'll be a massive time saver for rapid release development.
Not that we know for sure, but reporting that the board wanted to be not-for-profit should be a warning sign they don't see customers as the people (revenue) they need to protect and grow.
We've seen it many times. Start-ups run to capture the whales of the market, who are more than happy to screw smaller customers or bate and switch free users.
https://www.bloomberg.com/opinion/articles/2023-11-20/who-controls-openai
Much of the superstardom of these people is due to the work being done in public with a lot of experimentation going on.
So what to do if you are Apple's AI people? If any good then likely they may defect or renegotiate packages.
The whole event may catalyse side tremors in other businesses like Apple as people re-evaluate the landscape.
MS is now doing their own server side AI silicon, Google has done it for years, AWS as well, Tesla is... Meta and others may as well?
Apple - where you at? Relying on enthusiasts to highlight how good client side silicon is can be useful very short term. Not a strategy though and also does not solve Siri is a green screen CLI.