Hacker News Reader

Comment by righthand | Hacker News Reader

Why? The board announced the removal of a CEO which they have the power to do. It upset a bunch of employees who liked the CEO. The board is in search of a new CEO.

OpenAI’s valuation isn’t because Sam Altman is worth $90B it’s the tech. The tech is fine. Investors have no reason to worry about their investment in the tech.

The people under pressure are Sam Altman trying to reclaim his job he was fired from and Satya Nadella who is running a campaign to eat OpenAI. None of that is related to the functioning tech of OpenAI, all of it is related to the drama of what happened. Drama is not equal to money.

I was sad when my favorite pet died too, but then eventually you get a new pet that’s just as good as a Worldcoin guy, maybe better.

View on HN

> The board announced the removal of a CEO which they have the power to do.

They have the power to remove him momentarily, do they have the power to keep him out? we're still finding out.

> OpenAI’s valuation isn’t because Sam Altman is worth $90B it’s the tech.

OpenAI's valuation is nominally because of the tech. It's fundamentally because of the people. 90% of the staff is threatening to follow Sam to $MS so that's the whole ball game. They went to OpenAI to make themselves rich. The board fired the guy that was trying to do that for them.

> I was sad when my favorite pet died too, but then eventually you get a new pet that’s just as good as a Worldcoin guy, maybe better.

Doesn't matter if you or I am sad. It only matters if the talent at OpenAI is sad. The board made them sad. Sam and Sataya make them happy.

If it’s fundamentally about the people, can you name some of these 90% that OpenAI couldn’t live without? Other than Ilya? From where I’m sitting the 90% are replaceable in a flooded market of available software engineers.

If Satya and Sam make them happy then go work for Microsoft. The board is calling their bluff right now.

Do you think they hired those people because they're useless?

> The board is calling their bluff right now

The board is not doing much of anything. They picked a CEO then she went to team sam. Then they picked another CEO and now he's threatening them with resignation. Not to mention the 3 other people they tried to bring into the role who refused.

No I think they hired them because they can do the job the agreed to pay them. I don’t think these people are irreplaceable, that would be silly. That’s what Sam Altman wants you to think.

> I don’t think these people are irreplaceable, that would be silly.

Maybe go and spend some time to bone up on who exactly works for OpenAI, what their track record is (education, publications and so on) and then tell me where you are going to find another 700 people that talented who will want to work for a company this damaged.

That’s only if 700 people are all doing the same job. The entire company is key engineering talent? That seems unlikely.

> From where I’m sitting the 90% are replaceable in a flooded market of available software engineers.

The market is flooded with junior engineers who have recently graduated from coding boot camps, and who might be good at building the specific types of software they have been shown. However, there is, as I understand it, still a dearth of senior engineers who can do the sort of work that OpenAI engineers are doing: building a novel product, with novel, cutting edge algorithms, at massive scale.

If lots of people could easily replicate OpenAI’s product, it seems strange that has nobody released a competitor to gpt4. Why is Microsoft - a company with one of the largest pools of software engineers in the world - clambering to hire all the engineers at OpenAI?

It’s because those people aren’t replaceable. There is probably a global pool of maybe ten thousand people who can do the work that OpenAI is doing. And OpenAI has assembled a high performance team of ~800 of them. Those people are currently worth their weight in gold.

It is not flooded with junior engineers, it's engineers of all skill levels. I happen to know as someone who is senior looking for work right now. Most of the open positions are for senior and staff level positions.

Microsoft isn't clambering to hire them, they're clambering for the IP. The OpenAI employees will get significantly less if they go over to Microsoft.

My guy. It’s flooded with juniors. Juniors at all levels of junior-ness. Including ones called “senior”

I hope you enjoy the difficulty of finding a job in a flooded market one day.

I hope no one does. Whether it's the Open AI situation or the job market, I am describing reality, not wishful thinking.

> Most of the open positions are for senior and staff level positions.

So what you're saying is that companies are struggling to find qualified senior and staff level engineers to fill their open roles? Doesn't sound like a flooded market to me.

I can't name them, no, but they're the people who built the thing that is being valued 90B. If they migrate (leadership included), so does the potential for future value. My feeling is that the 90B are moving. Sure, IP, existing contracts, licensing, you name it, will have residual value - but if I consider its board a clown car I will refrain from building on top of its API because they might not be around by lunchtime tomorrow. Enough people think like this, it starts spiraling down the drain.

Do you know the board is a clown car or is that what the media surrounding “the chaos” has told you. The board has been quiet is all. Does that automatically mean they’re in the back flushing documents down the toilet and juggling balls? Or is there a better reason they feel no reason to publicly respond to a bunch of media flurry?

https://twitter.com/OpenAI/status/1727206187077370115

Sam and Greg returning. Board folded.

https://twitter.com/emilychangtv/status/1727228431396704557

Their hand picked CEO called them on their BS and they could not produce

About the 90%, this is very much a game of musical chairs and game theory in general. If your perception is that there is a high likelihood that Altman might return it's prudent to sign the letter. That's why I think the past few days were about shaping a narrative on Twitter and in the media in general.

> OpenAI’s valuation isn’t because Sam Altman is worth $90B it’s the tech. The tech is fine. Investors have no reason to worry about their investment in the tech.

The following story is only about an hour old, but the investors are reportedly considering legal action against the board:

https://www.reuters.com/technology/openai-investors-consider...

Technology is not made out of thin air but it is made - and maintained - by people; and when 500 out of 700 people threaten to quit on the spot and leave for competitors, or for Microsoft, investors might not be happy at all.

It's not 500/700. It's like 728/770 last I checked. Almost everyone.

Yes but as mentioned most of these people can be replaced.

> OpenAI’s valuation isn’t because Sam Altman is worth $90B it’s the tech.

The IP might be worth something (no idea what kind of patents they hold), but the valuation is fundamentally based on the current products they have and promise of future tech the company can create. And the future value stream far outweighs the current product.

And none of the future product is going to happen without people.

Of course, OpenAI without Sam will hire new people. But if they have already hired the best and they already had worthwhile experience, hiring new people has almost zero chance to recreate their ability to create new products at least in short/medium term.

Finding the right people and hiring them is far more difficult than it is assumed. Dropping money on the problem does not solve more than half of it.

"It upset a bunch of employees. The tech is fine". Uhmm, it upset 98% (not figuratively, but mathematically) of all 700 employees who signed a petition to quit unless he's reinstated and for the board to resign. As for the tech... who... exactly do you think created, maintains and is developing future enhancement for the tech? Is it doing it by itself? Sorry, we're not there yet.

> Uhmm, it upset 98% (not figuratively, but mathematically) of all 700 employees who signed a petition to quit unless he's reinstated and for the board to resign.

And they have yet to leave as far as I know. Seems like no one is in an actual hurry to quit.

Sam Altman is currently talking to the board. You want them to spin up drama by quitting in the middle of the talks? You're trying too hard to copy paste your desired narrative on this and it fits so, so, so poorly.

If Sam comes back it’ll be under heavy changes. The board isn’t going to leave. You don’t get ousted and then get to call the shots, no matter how many employees you think you have on your side. Furthermore the fact that Sam is in talks is further proof that he won’t be coming back on his terms. As the board why would you fold to the employees just because they like a guy?

I’m not copy pasting anything, cute diss though.

Sorry to be blunt, but literally everything you said so far was nonsense, so I'll understand your opinion as a sign that Sam Altman will come back and the board will resign.

Ah I see more disses, very cute. Get a Team Sam t-shirt made for yourself.

Another bad guess from you.

Oh and:

https://twitter.com/OpenAI/status/1727206187077370115

"We have reached an agreement in principle for Sam Altman to return to OpenAI as CEO with a new initial board of Bret Taylor (Chair), Larry Summers, and Adam D'Angelo.

We are collaborating to figure out the details. Thank you so much for your patience through this."

Get yourself a "loser" shirt.

To be fair if this board is good at one thing it is spinning up drama

I think one lesson we learned from Twitter is that you can cull a large amount of a tech workforce and the product will keep humming along, albeit at a reduced rate of change.

Also, unlike Twitter, OpenAI pays engineers very competitive salaries and can aggressively re-hire roles.

You learned the wrong lesson. Twitter is dying and operating at a significantly degraded level, including moderation, content filtering, ad positioning, which caused a third or fourth mass advertiser exodus just few days ago.

Also. Leading the world in revolutionary AI is not exactly like maintaining a pre-existing message app.

Twitter:

- Had 7500 employees before Musk fired most of them.

- Now 1300, of which 550 are full-time devs.

- That's a company valuated at less than $16 billion currently, and losing about $2 billion a year.

OpenAI:

- Has 700 employees, including world-leading scientists and researchers that are practically irreplaceable.

- If they resign, it'll have 14 employees.

- For a company that before Sam Altman was fired was valuated at $90 billion.

I dont think there is a large overlap here.

Twitter seemed to be adult daycare with a HUGE amount of positions that openly did nothing all day, many many positing publicly about how they only actually worked a few hours a week. Along with a Huge business support staff that seemed to really be pointless

I do not see that analog at OpenAI which has mainly engineering positions.

Also the idea that you just just "aggressively re-hire role" engineering positions in a emergent field like AI seems to be bit naive to me. How many experience people outside OpenAI are their int he first place??

It seems to me to be very ironic that in other situations is seems to be common for people to hate on corporations that treat their employees like interchangeable widgets that can just be replaced on a dime with a new human, yet here we see people claim that exact thing. Very odd

There is no comparison between the cadre at Twitter and and the talent pool at OpenAI, I don't think you can replace the bulk of them at all, and nobody worth that kind of compensation is going to touch OpenAI with a 100 foot pole as long as the current board is there.

Twitter is a very different animal. A twitter clone can be thrown together over a weekend with average web stack coders, making it scalable is a bit more work true but nothing fundamentally new or revolutionary. OpenAI on the other hand is working on new, challenging hard problems in a complex feild of math/computer science.

98% is meaningless. If Altman is reinstated then he will have the power to fire anyone who didn’t sign. You’d be an idiot not to sign it.

If they didn't sign he wouldnt have the leverage to come back in the first place

Sure if you’re one of the first names to sign. But let’s say most of the company has signed, you earn a 19-bit salary, there’s now a good chance that this guy will come back and be your boss again. Are you really going to refuse to sign and withhold your support for him? No way.

According to Stratechery, Microsoft does not just have the rights to use the source code and model weights for GPT-3.5 and GPT-4, but a clause saying that Microsoft could not use the models as foundations for their own projects was recently removed.

> Investors have no reason to worry about their investment in the tech.

If Stratechery is correct, this means that if Microsoft successfully poaches Sam Altman and friends, and vacuums OpenAI clean of talent, they will even be able to use the actual GPT-4 to build a Microsoft GPT-5. That should make investors sweat. If you lose the people and (possibly) the technology, what's left?

Source? Ben didn't read his cited source in his article yesterday.

Currently, the claim is unproven. However, if I were an investor, I'd be pounding on the doors demanding for clarification.

I meant, can you link the claim that the clause in question got removed?

https://stratechery.com/2023/openais-misalignment-and-micros...

"Microsoft’s original agreement with OpenAI also barred Microsoft from pursuing AGI based on OpenAI tech on its own; my understanding is that this clause was removed in the most recent agreement."

(It's hidden under a citation bubble you need to click after the third paragraph.)

I see. Pursuing AGI sounds distinct from commercializing future versions of ChatGPT based on OpenAI IP. I imagine the exact legal wording in the contract makes all the difference.

The WSJ article from June he cited indicates the license still comes with restrictions.

https://www.wsj.com/articles/microsoft-and-openai-forge-awkw...

I'd have my lawyers do the pounding.

Yeah okay Microsoft poaches them but they still have to restart building everything from scratch which wouldn’t benefit Microsoft. Microsoft doesn’t want to poach talent, they can always do that with or without this incident. Microsoft wants the whole bundle. Hence the media campaign Satya is doing.

No, the point is they would not have to start building everything from scratch.

That only happens if Microsoft acquires all of OpenAI, not if they just poach the employees. Poaching means starting again even if they have the road map in their head.

They already have access to GPT4 per their agreement with OpenAI.

They claimed to remove the CEO with cause. They have not provided cause. Even the twitch dude is asking for the cause.

The board just firing off half-cocked and then going silent is probably a larger source the issues for employees than altman the person/leader. There is effectively nobody in the driver seat and the board can't be trusted to do anything in the best interests of the company.

They did provide cause, you just don't accept what they said as valid or detailed enough.

No they made an extremely vague aligation with no concrete examples or evidence provided.

Is it the tech that is worth that much or the tech and the people that built the tech when future iterations of the product rely so heavily on the people? Losing 95 percent of the employees will definitely impact delivering future products and that will definitely impact future profits and that has to impact valuation.

> The tech is fine.

Have you ever maintained the software where the original team left? I have. And though it was much less complex and cutting edge than what OpenAI is doing, it was a massive pain with loss of knowledge and time.

Yes I have. I’ve even added features to legacy software and have made that a part of my mentoring to engineers to not reinvent the wheel and be good at reading and understanding code. I don’t think it’s a loss because the company has chosen to invest in that software. I include that time as onboarding for a project.

OpenAI IP is not code that you can read. It's the know-how of their team.

The board tried to get a merger with Anthropic, probably the second best team in the world for this type of work.

They don't seem to think they can "just re-hire".

It's not about what the board can or can not do. It's about whether they can make it stick. I'm thinking they can't but plenty of clever people here on HN think they can. Time will tell.

Vindicated. I've never had such pushback on something that seemed so obvious.

> OpenAI’s valuation isn’t because Sam Altman is worth $90B it’s the tech.

It's the current AND future tech. If most of the people who made that tech jump ship, that does not bode well for the future development of the technology. Maybe if the tech were something more ordinary, sure, but I think we both agree that it's not. For an investor, the future prospects of the company are generally a lot more important than its present state.

Companies aren’t valued for their current state, but future prospects.

Is Sam Altman a future prospect of LLMs? It’s not like he holds the secret to GPT v5 or anything.

I'd be surprised if Sam was by himself, but I wouldn't be at all surprised if some of the people who signed that letter do hold the secret to the next big thing (GPT-5 probably isn't the next big thing; if it exists already it's likely just a nice iterative improvement and not a fundamental change).

Seeing the amount of employees that are willing to follow him, yes I'd say he is.

If you have a mass exodus of employees, sure you can hire people to maintain ChatGPT 4, but they won't complete a GPT 5 or 6. That valuation is definitely based on the assumption that they will continue advancing AI and that they will consistently stay ahead of competitors for some time. They just axed both of those assumptions.

ChatGPT 4 as the endgame is not worth 90 billion dollars.

> Investors have no reason to worry about their investment in the tech

What? They have a MASSIVE reason to worry: 3/4 board members are clearly not aligned with the profit motive. And D'angelo has a massive conflict of interest with monetizing their models.

How are they not aligned? They didn’t oust Sam because of profit they ousted him for dishonesty.

are you serious? they're board members of the non-profit... they don't have any shares in the for-profit subsidiary and thus no financial incentive for commercializing the IP

meanwhile, all the employees are paid via a profit sharing scheme. 98% of them are backing Altman because they don't want to see their compensation nuked.

How does ousting Altman nuke compensation? If they keep working there and bring in a new CEO equivalent to Altman with less dishonesty then everyone still wins, except Altman.

Reply on news.ycombinator.com