Really part of another era, where AT&T sunk billions into the labs just for pure research. If only corporations could be a far seeing today.
Related acecdote: Charles Townes described that after the invention of the maser, he had to make a case for Bell Labs to patent it, against some resistance. As he described it, the eventual winning phrase was: "Well, it could be used for telecomunication".
FAANG do a lot of research, but the budgeting for those efforts is much more tightly controlled and connected to corporate strategy.
My hope is that we can establish a similar model for biology, since many challenges in that domain are pre-commercial but not at the appropriate scale for academic labs.
The more substantial point is breadth, though. AT&T research at its peak was more comparable to a university than an industrial R&D institution, with both basic and applied research occurring in a huge variety of fields---likely all fields of physical science at various points. All of the organizations you mention are essentially single-purpose. Some of the major tech companies (Google, Microsoft) are known for wide-ranging R&D, but they seem to have struggled to produce the kind of basic advances that AT&T did, and also seem to be backing off of that work.
Basic research is the big problem... AT&T had very little problem with money going to basic research without known applications, likely due to the substantial influence that academia and "scientific management" had on AT&T, as well as its reputation for excellent research management practices that made AT&T a major part of basic physics research on the behalf of the DoD for decades (Sandia National Laboratories, for example, operated by AT&T until relatively recently). In my opinion, accounts of Bell Labs enormous success, having made major contributions to nearly every field of engineering, often focus too much on the excellence of its staff (which was a factor, although Bell Labs management practices were not exactly modern), and downplay the role of AT&T's willingness to fund basic research and ability to find outside funding for basic research as well. Few modern organizations seem to be willing to entertain major efforts in a nascent field of astronomy, for example, on the assumption that the technical work will lead to useful engineering knowledge.
There's a lot out there about the history of Bell Labs, but its own multi-volume "A History of Engineering and Science in the Bell System," from the late '70s, does a good job of covering the incredible breadth and scale of work performed at Bell Labs. One thing to consider is that costs were quite a bit lower back then, and so Bell Labs small-looking budget of $3-5 billion of today's dollars per year translated to a workforce in the tens of thousands working, during WWII, 70-hour weeks. A two-day weekend did finally reach most of the Bell Labs technical staff early in the Cold War.
Besides, if you read into the history of Bell Labs, you will recognize vestiges of its bureaucratic structure in most government and military R&D today. Its influence in government and military physical research is enormous.
disclaimer - i work there as a senior researcher and we still do world class research, so pls don't portrait us as a thing of the past just yet, thx :)
We can probably get such things back, but they are not Free exactly.
RCA had similar labs, but I think changes to the tax code makes them less worthwhile to do today as well.
I think it's only fair to mention that AT&T's motives here weren't entirely selfless: the value of much of their research was really only realized by the public as a result of a 1956 consent decree. Before that, it was largely locked behind patents; the consent decree forced them to license their research without royalties (and forbade them from competing in non-telecommuncations domains).