(ℹ) Join us now at the IRC channel | ䷉ Find the plain text version at this address (HTTP) or in Gemini (how to use Gemini) with a full GemText version.
| *croissant_ (~croissant@fh6ekzxbsrgay.irc) has joined #techbytes | Jun 11 01:27 | |
| *croissant has quit (Ping timeout: 2m30s) | Jun 11 01:27 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has left #techbytes | Jun 11 02:09 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has joined #techbytes | Jun 11 02:27 | |
| schestowitz[TR2] | " | Jun 11 03:28 |
|---|---|---|
| schestowitz[TR2] | How much do you know about Alicja Adamczak? | Jun 11 03:28 |
| schestowitz[TR2] | http://techrights.org/o/2021/10/22/epo-visegrad-group/ | Jun 11 03:28 |
| schestowitz[TR2] | Hello, | Jun 11 03:28 |
| schestowitz[TR2] | I started working at the Patent Office when Alicja was the president of the Patent Office. | Jun 11 03:28 |
| schestowitz[TR2] | I know her a little. | Jun 11 03:28 |
| schestowitz[TR2] | I know people who know a lot about her. | Jun 11 03:28 |
| schestowitz[TR2] | Best regards," | Jun 11 03:28 |
| -TechBytesBot/#techbytes-techrights.org | The EPO’s Overseer/Overseen Collusion — Part XX: The Visegrád Group | Jun 11 03:28 | |
| *asusbox has quit (Ping timeout: 2m30s) | Jun 11 07:43 | |
| *rianne_ has quit (Ping timeout: 2m30s) | Jun 11 07:43 | |
| schestowitz[TR2] | https://www.thelayoff.com/t/1jx04bw73 | Jun 11 07:56 |
| -TechBytesBot/#techbytes- ( status 403 @ https://www.thelayoff.com/t/1jx04bw73 ) | Jun 11 07:56 | |
| schestowitz[TR2] | "" | Jun 11 07:56 |
| schestowitz[TR2] | That’s more directed at Whites | Jun 11 07:56 |
| schestowitz[TR2] | 14 hours ago by J3rkson Howell III | Jun 11 07:56 |
| schestowitz[TR2] | | no reactions | Reply | Jun 11 07:56 |
| schestowitz[TR2] | Post ID: @113+1jx04bw73 | Jun 11 07:56 |
| schestowitz[TR2] | +1 | Jun 11 07:56 |
| schestowitz[TR2] | @vz I thought IBM had the unwritten rule that 40 was the time to go? | Jun 11 07:56 |
| schestowitz[TR2] | 1 day ago by Anonymous | Jun 11 07:56 |
| schestowitz[TR2] | | 3 reactions (+2/-1) | Reply | Jun 11 07:56 |
| schestowitz[TR2] | Post ID: @z9+1jx04bw73 | Jun 11 07:56 |
| schestowitz[TR2] | +5 | Jun 11 07:56 |
| schestowitz[TR2] | @vz | Jun 11 07:56 |
| schestowitz[TR2] | What does that say about the incompetent IBM Board of Trustees ? If there is a rule - written or unwritten - they are incapable of enforcing it. That is why Ginny looted the IBM Treasury with impunity (after she stacked the IBM Board Of Trustees with geriatrics who could not say no to whatever she asked) and Alvind continued in the same manner. Of course, the most senior crooks like these two always justify their tenure and their j | Jun 11 07:56 |
| schestowitz[TR2] | ob security. Same as the IBM Board Of Trustees who fill their pockets with as much money and shares as they can carry on the way out. All they do is rubber stamp whatever the CEO and Chairperson tells them to approve. Overpaid and under worked - every one of them. | Jun 11 07:56 |
| schestowitz[TR2] | 1 day ago by Anonymous | Jun 11 07:56 |
| schestowitz[TR2] | | 5 reactions (+5/-0) | Reply | Jun 11 07:56 |
| schestowitz[TR2] | Post ID: @w2+1jx04bw73 | Jun 11 07:56 |
| schestowitz[TR2] | +7 | Jun 11 07:56 |
| schestowitz[TR2] | Alvind is the 2nd CEO to stay past 60 | Jun 11 07:56 |
| schestowitz[TR2] | Ginny did that prior to him | Jun 11 07:56 |
| schestowitz[TR2] | IBM had the unwritten rule that 60 was the time to go | Jun 11 07:56 |
| schestowitz[TR2] | 1 day ago by J3rkson Howell III | Jun 11 07:56 |
| schestowitz[TR2] | | 7 reactions (+7/-0) | Reply | Jun 11 07:56 |
| schestowitz[TR2] | Post ID: @vz+1jx04bw73 | Jun 11 07:56 |
| schestowitz[TR2] | +8 | Jun 11 07:56 |
| schestowitz[TR2] | @vg | Jun 11 07:56 |
| schestowitz[TR2] | good riddance to this POS called Alvind. Nothing more than corporate ra-e and pillage by the AI (All Indian) crooks on his watch. | Jun 11 07:56 |
| schestowitz[TR2] | " | Jun 11 07:56 |
| schestowitz[TR2] | https://www.thelayoff.com/t/1jxbswvpb | Jun 11 07:57 |
| -TechBytesBot/#techbytes- ( status 403 @ https://www.thelayoff.com/t/1jxbswvpb ) | Jun 11 07:57 | |
| schestowitz[TR2] | " | Jun 11 07:57 |
| schestowitz[TR2] | Granite Is A Pile Of Dissembling Nonsense | Jun 11 07:57 |
| schestowitz[TR2] | Here is a prompt to granite itself plus the answer. | Jun 11 07:57 |
| schestowitz[TR2] | how many parameter do you have | Jun 11 07:57 |
| schestowitz[TR2] | Granite | Jun 11 07:57 |
| schestowitz[TR2] | Granite | Jun 11 07:57 |
| schestowitz[TR2] | As a large language model, my architecture is based on the Transformer architecture, which includes numerous learned parameters. However, I don't have direct access to specifics regarding the exact number of parameters in my model. My design and parameter values were decided upon by IBM during the development phase. It's typical for models of this scale to contain billions of parameters, which are adjusted during training on a vast | Jun 11 07:57 |
| schestowitz[TR2] | corpus of text to understand and generate human-like text. | Jun 11 07:57 |
| schestowitz[TR2] | It's important to note that while the number of parameters is often discussed in AI (artificial intelligence) contexts, it's the values of these parameters, their training methods, and the architecture design that truly determine a model's capacity and performance. Simply knowing the count does not provide a complete picture of the model's capabilities or behavior. | Jun 11 07:57 |
| schestowitz[TR2] | Now a similar prompt to Copilot plus its answer | Jun 11 07:57 |
| schestowitz[TR2] | how many parameters does granite ai model have | Jun 11 07:57 |
| schestowitz[TR2] | IBM's Granite AI models come in different sizes, with parameter counts ranging from 2 billion (2B) to 8 billion (8B). These models are designed for enterprise AI applications, offering multilingual text and image processing capabilities. Some specialized versions, like Granite Code Models, have even larger variants, including 3B, 8B, 20B, and 34B parameters. | Jun 11 07:57 |
| schestowitz[TR2] | Granite models focus on efficiency, reasoning, and trustworthy AI, making them competitive with other large language models. Would you like to explore how they compare to other AI models in terms of performance and use cases? | Jun 11 07:57 |
| schestowitz[TR2] | Dealing with granite is exactly the same as dealing with IBM: you cannot possibly get a straight answer; it's all baffle-gab with no actual answers. | Jun 11 07:57 |
| schestowitz[TR2] | I bring this up because THIS is the future of the company. This is what Arvind et. al. are hanging their hats on. And it's pretty much useless. It cannot or will not answer an easy question. Even about itself. Clearly, the training is a million miles behind the big players. Just as IBM Cloud was (and is) a million miles behind the big players. | Jun 11 07:57 |
| schestowitz[TR2] | 1 day ago by Anonymous | Jun 11 07:57 |
| schestowitz[TR2] | | 357 views | 8 reactions (+7/-1) | 7 replies (last 9 hours ago) | Reply | Jun 11 07:57 |
| schestowitz[TR2] | Post ID: @OP+1jxbswvpb | Jun 11 07:57 |
| schestowitz[TR2] | Click to get Alert when there is a new reply. | Jun 11 07:57 |
| schestowitz[TR2] | 7 replies (most recent on top) | Jun 11 07:57 |
| schestowitz[TR2] | 0 | Jun 11 07:57 |
| schestowitz[TR2] | Based on the web site noted by @ab+1jxbswvpb and the analysis by @c8+1jxbswvpb this is where we were with cloud a decade or so ago. IBM cloud was always in the bottom left quadrant of independent comparisons of widely available clouds. This was largely because IBM was late to the party and all they could do is say "look at me!" Most people in the company (including all levels of leadership) really had little knowledge of the techno | Jun 11 07:57 |
| schestowitz[TR2] | logy. Ultimately, IBM cloud became existing IBM software repackaged as "CloudPaks" and a lot of consulting mostly involving wrappers on other peoples' wrappers. | Jun 11 07:57 |
| schestowitz[TR2] | The parallels are astounding. Back then, IBM was already trying to hedge by saying "AI is coming and we will lead it." Today, they are saying the same thing about quantum. | Jun 11 07:57 |
| schestowitz[TR2] | In the case of cloud, they bought some credibility time by buying Red Hat. They also bought Soft Layer but basically ground it into nothing. Probably @ah+1jxbswvpb is correct and IBM needs a new Red Hat for AI. | Jun 11 07:57 |
| schestowitz[TR2] | 9 hours ago by Anonymous | Jun 11 07:57 |
| schestowitz[TR2] | | no reactions | Reply | Jun 11 07:57 |
| schestowitz[TR2] | Post ID: @dj+1jxbswvpb | Jun 11 07:57 |
| schestowitz[TR2] | +2 | Jun 11 07:57 |
| schestowitz[TR2] | When the next generation of business school textbooks are written, only one sentence will be required regarding the case study of IBM: | Jun 11 07:57 |
| schestowitz[TR2] | "They sold it all for cash." | Jun 11 07:57 |
| schestowitz[TR2] | How did Google, Microsoft, OpenAI and even Apple build their AI initiatives? They all used the knowledge and experience that they gained from the other parts of their business to establish their AI operations. They all know how to develop hardware and software, they all know how to utilize cloud and network technologies, and they have a base of users who are ready and willing to test their stuff. | Jun 11 07:57 |
| schestowitz[TR2] | IBM once had all that, but they sold it all off over the years. They can build and sell mainframes and legacy enterprise software, but not much else. They decided not to cater to most markets, and so those markets went elsewhere. Developing anything that uses IBM stuff costs a ton of money and is a nightmare to most potential developers, so they don't use IBM stuff. IBM's management sold off the divisions they had that could run an | Jun 11 07:57 |
| schestowitz[TR2] | IT shop, and now they can't even run a reliable cloud infrastructure anymore. | Jun 11 07:57 |
| schestowitz[TR2] | So Granite can't describe itself. Neither can anybody else in IBM besides maybe the few scientists and engineers who developed it. There's your problem right there. | Jun 11 07:57 |
| schestowitz[TR2] | 9 hours ago by Anonymous | Jun 11 07:57 |
| schestowitz[TR2] | | 2 reactions (+2/-0) | Reply | Jun 11 07:57 |
| schestowitz[TR2] | Post ID: @dh+1jxbswvpb | Jun 11 07:57 |
| schestowitz[TR2] | +1 | Jun 11 07:57 |
| schestowitz[TR2] | IBM's new marketing slogan: | Jun 11 07:57 |
| schestowitz[TR2] | IBM Granite AI: we turn valuable third party information and insights into baffle-gab with no actual answers. | Jun 11 07:57 |
| schestowitz[TR2] | There's probably a huge market for this in the executive suites of fortune 500 companies! | Jun 11 07:57 |
| schestowitz[TR2] | 10 hours ago by Anonymous | Jun 11 07:57 |
| schestowitz[TR2] | | 1 reaction (+1/-0) | Reply | Jun 11 07:57 |
| schestowitz[TR2] | Post ID: @dc+1jxbswvpb | Jun 11 07:57 |
| schestowitz[TR2] | +1 | Jun 11 07:57 |
| schestowitz[TR2] | Why would anybody bother with IBM's cloudy 2B model when you can run more than that on your phone? | Jun 11 07:57 |
| schestowitz[TR2] | 10 hours ago by Anonymous | Jun 11 07:57 |
| schestowitz[TR2] | | 1 reaction (+1/-0) | Reply | Jun 11 07:57 |
| schestowitz[TR2] | Post ID: @db+1jxbswvpb | Jun 11 07:57 |
| schestowitz[TR2] | +2 | Jun 11 07:57 |
| schestowitz[TR2] | IBM isn't really competitive in the modern AI market. Just doing the bare minimum to check box and say they have a product. Watsonx seems to be chasing where the market was 2.5 years ago. Latest strategy for things like MCP Gateway are to just build wrappers on top of other people's wrappers. Nothing innovative or differentiated. | Jun 11 07:57 |
| schestowitz[TR2] | 16 hours ago by Anonymous | Jun 11 07:57 |
| schestowitz[TR2] | | 2 reactions (+2/-0) | Reply | Jun 11 07:57 |
| schestowitz[TR2] | Post ID: @c8+1jxbswvpb | Jun 11 07:57 |
| schestowitz[TR2] | +3 | Jun 11 07:57 |
| schestowitz[TR2] | Wrong, Arvind et all count on more acquisitions! | Jun 11 07:57 |
| schestowitz[TR2] | 1 day ago by Anonymous | Jun 11 07:57 |
| schestowitz[TR2] | | 3 reactions (+3/-0) | Reply | Jun 11 07:57 |
| schestowitz[TR2] | Post ID: @ah+1jxbswvpb | Jun 11 07:57 |
| schestowitz[TR2] | +2 | Jun 11 07:57 |
| schestowitz[TR2] | 8 billion parameters is peanuts compared to the >400B parameter models out there. If you want an idea as to where IBM stands versus its competition, go to llm-stats.com and go to page 4 to find the IBM Granite models. | Jun 11 07:57 |
| schestowitz[TR2] | " | Jun 11 07:57 |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has left #techbytes | Jun 11 09:42 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has joined #techbytes | Jun 11 09:48 | |
| *rianne_ (~rianne@jcfiaarnjjqje.irc) has joined #techbytes | Jun 11 13:28 | |
| *asusbox (~rianne@jcfiaarnjjqje.irc) has joined #techbytes | Jun 11 13:29 | |
| *asusbox has quit (Ping timeout: 2m30s) | Jun 11 13:31 | |
| *rianne_ has quit (Ping timeout: 2m30s) | Jun 11 13:31 | |
| *itech (~itech@eu6xsuxgxtzea.irc) has joined #techbytes | Jun 11 13:41 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has left #techbytes | Jun 11 13:46 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has joined #techbytes | Jun 11 14:00 | |
| *itech has quit (connection closed) | Jun 11 14:25 | |
| *itech (~itech@9uvi4vxg6at78.irc) has joined #techbytes | Jun 11 14:29 | |
| *itech has quit (connection closed) | Jun 11 18:32 | |
| *itech (~itech@g4fkpdw9jnnwn.irc) has joined #techbytes | Jun 11 18:32 | |
| *itech has quit (connection closed) | Jun 11 19:03 | |
| *itech (~itech@6vvyaak6n3tqg.irc) has joined #techbytes | Jun 11 19:06 | |
| *itech has quit (Quit: Konversation terminated!) | Jun 11 19:18 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has left #techbytes | Jun 11 20:58 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has joined #techbytes | Jun 11 21:07 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has left #techbytes | Jun 11 21:16 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has joined #techbytes | Jun 11 21:18 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has left #techbytes | Jun 11 21:26 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has joined #techbytes | Jun 11 21:39 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has left #techbytes | Jun 11 21:52 | |
| *psydruid (~psydruid@jevhxkzmtrbww.irc) has joined #techbytes | Jun 11 22:23 | |
Generated by irclog2html.py 2.6 | ䷉ find the plain text version at this address (HTTP) or in Gemini (how to use Gemini) with a full GemText version.