IBM to Acquire Confluent

Surely by now everyone, including non-developers and non-software people, know exactly what IBM is, and you don't sell to IBM/join IBM without knowing exactly what's about to happen. No one joins IBM today and thinks there will be a huge focus on customer satisfaction or focus on great product design, it's all about squeezing maximum profit out of products until you need to discontinue them because you chased away all of the customers.

    > ...and internal systems they have to use, whose sole purpose is to make sure nobody does anything

In defense of Byzantines. Their rules and amazing diplomatic prowess is what let them be an empire for so long. The negative connotations to Byzantine comes from the negative perception the west had of them. Byzantines were very practical in regards to who they allied with.

I hope Hashicorp survives. A few higher ups I’ve talked to there made it seem like IBM wants to learn from them, not force their old ways onto Hashicorp. We’ll see. That one is still pretty new.

IBM is designed to milk every last bit of money from their clients. So they need to add new products every now and then to add new money flows.

Pretty bleak, and describes my experience to a T (although involving other companies). Has there ever been an example where a company has been acquired and culture/morale/conditions have actually improved rather than dissolved?

Why in the world would economists need to study this? It's been known that large bureaucracies have been dysfunctional for over a couple of decades now if not centuries. The large reason is because 1) the incentives to do great work are not there (most of the credit for a huge company's success goes to the CEO who gets 100X the salary of a regular worker while delivering usually pretty much nothing) 2) politics usually plays a huge role which gives a huge advantage to your competition (i.e. your competition needs to spend less time on politics and more time on the actual product) and 3) human beings don't functionally work well in groups larger than 100-250 due to the overwhelming complexity of the communication needed in order to make this type of structure work. Incentives though I think are the primary driver - most people at companies like IBM don't have any incentives to actually care about the product they produce and that's the secret behind the ruin of almost every large company.

Edit: you also seem to be giving too much credence to Watson. Watson was actually mostly a marketing tool designed to win in Jeopardy and nothing else. It was constructed specifically to compete in that use-case and was nowhere near to the architecture of a general transformer which is capable of figuring out meta-patterns within language and structurally understanding language. You can read about Watson's design and architecture here if you're curious: https://www.cs.cornell.edu/courses/cs4740/2011sp/papers/AIMa...

>Why do they keep doing the same thing again and again? How are they generating actual revenue this way?

IBM has a ton of Enterprise software, backed by a bunch of consultants hiding in boring businesses/governments.

They also do a ton of outsourcing work where they will be big enterprise IT support desk and various other functions. In fact, that side has gotten so big, IBM now has more employees in India in then any other country.

Your fascination seems hinged on the fact that IBM has "lots of talent to make ML/LLMs work" which judging by what they've put out so far and talk publicly about, is very far from the truth. Anyone who has a clue seems to (rightly) have left IBM decades ago, and left are business people who think "Managed to increase margin by 0.1%" is something to celebrate.

> Ok, so does anyone remember 'Watson'? It was the chatgpt before chatgpt. they built it in house

I do. I remember going to a chat once where they wanted to get people on-board in using it. It was 90 minutes of hot air. They "showed" how Watson worked and how to implement things, and I think every single person in the room knew they were full of it. Imagine we were all engineers and there were no questions at the end.

Comparing Watson to LLMs is like comparing a rock to an AIM-9 Sidewinder.

To add to that i think their R&D labs along with HPE were one of the few to innovate on the memristor and actually build some fascinating concept machines.If i rememeber HPE's was 'The Machine'.

Athough i think they just di/dont know how to adapt these to market that isnt a enterprise behemoth , rather than develop/price it so more devs can take a hold and experiment.

> Ok, so does anyone remember 'Watson'? It was the chatgpt before chatgpt. they built it in house. Why didn't they compete with OpenAI like Google and Anthropic are doing, with in-house tools?

Leadership in IBM also thought that Watson was like what what OAI/Anthropic/Google are doing now. It wasn't. Watson was essentially a ML pipeline over-optimized on Jeopardy, which is why it failed in literally every other domain.

Outside of Jeopardy, Watson was just a brand.

There are entire niches of us that make a living (not at IBM) making certain IBM products actually do what they're supposed to. From my vantage point I see essentially zero maintenance going on with their products. I sincerely don't understand the market (why do people keep paying hundreds of thousands to millions of dollars for non-existent support?) - but whatever.

I’m pretty convinced there is a bell curve of “understanding what IBM does” where idiots and geniuses both have absolutely no idea.

It really is probably that strangest company in tech which you think could be mysterious and intriguing. But no one cares. It’s like no one wants to look behind the boring suit and see wtf. From my low point on that bell curve I can’t see how they are even solvent.

Why didn't you ask to get the accounts provisioned?

We moved off HashiCorp's Terraform Cloud when they tried to hike the price 100x on us, although that was technically pre-acquisition I think (it was their move to resource-based pricing). In talking with our account manager, they basically said they only really cared about enterprise accounts, and that migrating away would probably make sense for us.

HashiCorp also changed their licenses to non-open-source licenses, but again I think this was technically pre-acquisition (I think as they were gearing up to be a more attractive target for an exit).

The argument has been made that the real value of RH lies in the people working there. And if IBM were to interfere too heavy-handedly, those people would just leave, and RH would become basically worthless.

Well, there is CentOS Stream:

https://www.centos.org/centos-stream/

And Fedora is still the upstream of RHEL, nothing changed there.

I migrated our company off Terraform to Pulumi as a direct result of the acquisition.

Gnome has stagnated significantly.

Every time an executive says AI the number goes up.

As I read the release, it just sounded like "something something something data, something something something AI."

AI is just the lastest buzzword. Everyone has it, because they have to. Don't look behind the curtain.

Event-driven AI decision making is the C-suite wet dream. A large % of major orgs run Kafka for their eventing systems.

They probably have cursor licenses

Streaming, EDA can solve lot of data challenges for enterprise AI use cases

We switched to Redpanda's BYOC product because we couldn't use Confluent Cloud (contractual reasons) and BYOC was a third the price of Confluent for Kubernetes while also being a managed service.

I've been pretty happy with RP performance/cost/functionality wise. It isn't Kafka though, it's a proprietary C++ rewrite that aims for 100% compatibility. This hasn't been an issue in the 2+ years since we migrated prod, but YMMV.

Maybe this whole thing it's because Snowflake acquired redpanda earlier this year: https://www.investors.com/news/technology/snowflake-stock-re...

I'm still convinced the vast majority of kafka implementations could be replaced with `SELECT * FROM mytable ORDER BY timestamp ASC`

Nothing wrong with Kafka. Time to build better abstractions on top of Kafka.

What are the alternatives?

Erlang/OTP!

wait what do you mean? what's wrong with kafka?

This isn't the old times, you can expect the opposite outcome these days.

I know companies who would certainly have fired people for buying IBM, if they could have gone back in time to do so.