Lina Khan points to Figma IPO as vindication of M&A scrutiny

The benefit is the same kind as from blocking a very small fraction of acquisitions. If you don't think the existing rules there are kafkaesque, then the small expansion shouldn't be a huge change. If you do, I'm not the person to argue with.

And these rules would very rarely come into effect.

If you don't understand why acquisitions might be blocked and the benefit of it then I'm not going to be the one to explain it.

I'm not here to defend the entire edifice of the FTC having control over acquisitions. I'm here to say that when they have that control, it should also include almost-but-not-quite-acquisitions.

I'm saying that what google did is effectively an acquisition except it broke the equity and it avoided regulation.

You keep making examples that are less clear-cut, but whatever I agree there are things a company can do that aren't effectively acquisitions. I'm not saying how to write the rule, I'm just saying google is on the wrong side.

> You are specifically saying that a company shouldn’t be allowed to license IP, shouldn’t be allowed to acquire the company and shouldn’t be allowed to hire the employees or some combination.

They're totally allowed to acquire the company... as long as the FTC doesn't say no. And they can do lots of those other things too.

I don't know how you got the impression that I think acquisitions should be blocked more than a fraction of a percent of the time.

Even further, google SHOULD have acquired the company instead of doing what they did.

> So in what world are startups going to have the funding or the infrastructure to have warehouses worldwide to compete with Amazon or servers to compete with AWS?

In a competitive market you don't have that. Instead of a massive conglomerate having a warehouse in every region, Alice, Bob and Carol each have a warehouse near New York, Dan, Erin and Frank have one near Houston, etc., and then a dozen independent aggregators each negotiate with a warehouse in each region to store goods for anyone who wants to offer fast delivery everywhere.

Meanwhile doing that, whether you're Amazon or not, is inefficient for anything that doesn't need to be delivered on short notice. If you have a recurring subscription to get a box of toiletries every month, it doesn't matter if it arrives on the 17th because they mailed it from a local warehouse on the 16th or a centralized warehouse on the 10th, and delivery companies offer discounts if you palletize shipments based on region even if they come from a central location, which removes the cost of having local warehouses for those regions.

> Or in Google’s case the infrastructure to design custom processors with the demand to actually buy enough slots from TSMC to make it affordable?

If there is aggregate demand for those processors then you sell them to the other people who want them regardless of whether they're within the same corporation, and then they don't have to be.

> You can’t make laws to undue scale efficiencies. A startup isn’t going to make a phone in their garage to compete with Apple no matter what magically thinking they have about the government passing laws.

The defect is in expecting one entity to make the entire phone.

One company makes a screen, one makes a battery, one designs a processor, another fabs it, another makes memory, another makes the OS (or it's open source), another lays out the system board to integrate the various components, another does final assembly, etc.

When you don't require the whole thing to be done under the same umbrella it doesn't take a trillion dollar company to do any given piece.

> Most startups can’t even pay the wages of a mid level employee at BigTech company that has been out of school for three years.

Suppose phone components were easily available as a fungible commodity, and had standardized interfaces so that integrating them was only a modest amount of work, i.e. a year of effort for a full-time engineer. Then you get a phone which sells as an also-ran -- a million units a year for three years, less than 0.5% of Apple's sales volume. The phone sells for $250, less than the cheapest new iPhone, and 0.5% of the retail price goes to pay the engineer.

Then they'd be making $3.75M for that year of work. Those numbers could be off by 10 fold and still be a competitive salary.

Except that the market is too concentrated and the standards don't exist, which means it's not that easy as things are now.

> Just like any other industry, if a startup can’t afford the free market price of labor, that’s a them problem.

The issue is, is it a free market, or a captured one?

The FTC caused this issue with over regulation and by creating an environment of fear that every acquisition is going to be scrutinized.

Google had no fiduciary duty to do what’s in the best interest of WindSurf’s investors.

Google had every right to give the WindSurf employees offers to leave. The employees had every right to accept the offers. There is no world where the government should have been involved in that part.

At that point, the Windsurf equity owners had two choices - either not license or license.

Which part should have been illegal? The FTC could have said “Google you can’t both get the employees and license the IP”. Google would have said “fine, we will just hire the talent and have them create something similar”. Windsurf would have been worse off.

Alternatively, Google could license the technology. Then the employees could say I would rather just work for Google. Would you then be okay with enforcing non competes even though they are currently illegal in California?

There is no law that you can come up with that doesn’t make it worse for both the employees and Windsurf.

We actually saw something like that happen between Microsoft and OpenAI during the Altman fiasco. Microsoft had already licensed the technology and Altman was going to go to Microsoft and take OpenAI’s best people with him. Should the government have stepped in then and told Microsoft they can’t hire OpenAI employees?