️️ ️️️️️️️️️️️ ️️ ️️ ️️ @moes_code

Exploring the frontiers of AI and LLMS to build, use and or deploy Machine Learning / AI solutions. moses-y.github.io 🇰🇪, Afrika Mashariki. Joined November 2017

Tweets

320K
Followers

7K
Following

8K
Likes

131K

️️ ️️️️️️️️️️️ ️️ ️️ ️️ @moes_code

8 hours ago

@chepkoech_bor I won't make it🥲

0 0 0 5 0

View Details

️️ ️️️️️️️️️️️ ️️ ️️ ️️ @moes_code

10 hours ago

Zee (Nktumalo) is one talented babe. Listen to her live kesho.

1 0 0 107 0

View Details

CJ Zafir @cjzafir

13 hours ago

I burned $3.40 on Jev in 24 hours. It's not Jev vs current LLMs. It'll be Jev + LLMs. Here are all the goods and bads of Jev. 1. Jev or a Jev-like system will be integrated into all LLMs for automatic model routing, small decisions, if/else queries, retrieval, tool selection, scoring, etc. For all these things, LLMs lose a lot of tokens on thinking and planning and use smarter instead of faster models. That can be fixed with Jev. 2. Jev can do things that require fine-tuning with LLMs. It can learn what to do from just 1 prompt. 3. It's crazy fast. It scans, analyzes, and gives its decisions instantly without a thinking blocker. 4. It doesn't waste tokens on thinking in plain English and planning. It scores probabilities and picks the best one. 5. Output cost of $0. Which makes sense because it's not built for free-form text generation like LLMs. It just outputs decisions. 6. It can fix RAG. The current 32k context is too low, but in the near future, we'll see systems like Jev with a 1M context window, and that'll fix the retrieval problem. 7. Jev has 3 decision types: multiple-choice questions, scoring the probability, and Noul. Choice → Which one? Score → How much? Noul → Yes or no? That's how it doesn't need reasoning. 8. Jev works on parallel sampling. LLMs generate sequentially: Token 1 ↓ Token 2 ↓ Token 3 ↓ Token 4 Jev can evaluate multiple questions against the same state in parallel. That's one of the reasons it has great latency. 9. Instead of RLHF (Reinforcement Learning from Human Feedback), it works on RLCD (Reinforcement Learning for Calibrated Decisions). With RLCD, the model's goal is not just to predict the answer; it also shows how much confidence it has. A model saying: true is different from: true, 99% confidence. 10. Jev response times are roughly: 70–500 ms, which is 193.6× faster and 444.6× cheaper than frontier LLM workflows. The biggest issue with LLMs today is that they take too much time to execute tasks. With a Jev-like system, that'll be at least 50% faster.

37 59 714 48K 634

View Details

Astrogato @astrocato

11 hours ago

Cuando fui a Bolivia y dije me gustan los felinos y me dijeron bueno entonces vas a cuidar 3 gatos monteses, un jaguar y un ocelote y tenía que sacarlos a pasear por el amazonas

58 1K 20K 345K 1K

View Details

️️ ️️️️️️️️️️️ ️️ ️️ ️️ @moes_code

8 hours ago

Tengu from 1300 years ago. How powerful is our 1000+ year old problem that's now in Palestine?

Holden_Onto_Truth @After_TheTruth

a day ago

I wonder who the Japanese artist was trying to sculpt?

70 1K 21K 265K 834

0 0 0 109 0

View Details

Darren Hope @DarrenH65352868

a day ago

@After_TheTruth Also the Japanese kappa another child snatching creatire from ancient Japanese folklore. That has a beak like nose and a circular area that could be mistaken for a little hat on the back of its head

1 11 179 6K 13

View Details

Bobby Axlerod @AxeCapitalNYC

a day ago

@After_TheTruth Tengu

6 33 475 8K 55

View Details

Holden_Onto_Truth @After_TheTruth

a day ago

I wonder who the Japanese artist was trying to sculpt?

70 1K 21K 265K 834

View Details

Rev. Poppy Haze 𓅋 @poppy_haze

17 hours ago

I love the discovery story too. An old grandpa finds what he thinks is an abandoned kitten on the side of the road and feeds it noodles and eggs for a year before realizing, “hey tilcayo, Youre not growing into a housecat, you’re just a little guy. I better tell Science”

National Geographic @NatGeo

a day ago

A new species of cat has been identified for the first time in a century—and experts think that there may be even more to discover around the world: on.natgeo.com/n5nxra

805 27K 158K 9.7M 14K

35 4K 59K 915K 4K

View Details

Elvis @elvissun

15 hours ago

Jev is INSANE. 🤯 in 24.9 seconds it read 384 news from this morning and told 15 brands which stories to hop onto today, for $0.19. Claude Opus 5, running on the same feed at the same time, got through 4/384 and cost $0.77. per headline that is ~390x cheaper, and the answer comes back before you finish reading the headline yourself. there's going to be so many ways for JEV to help you find trending stories, find journalists covering it, and get press coverage. everything in the video open-sourced here: newsjack.sh (in the demo/ folder, together with 30+ skills to turn your agent into a PR team)

109 185 3K 348K 4K

View Details

shirish @shiri_shh

14 hours ago

sir....we can't collect users' personal data ANYMORE they just launched a tool that swaps everything with FAKE DATA before it reaches us 😭

Peter Yared @peteryared

15 hours ago

Introducing AgentCloak: Use any AI without sharing your real data. Chinese AI services, ChatGPT, Claude, doesn't matter. You probably try to hide details before asking: different names, fake numbers, no address. But then the answer's useless because the AI is missing actual

268 786 5K 1.7M 6K

27 279 4K 526K 2K

View Details

Muffin 🇨🇿 @CatReviewer

a day ago

official behind the scenes footage

verz @nerdyverz

2 days ago

pose

76 5K 49K 2.0M 3K

86 11K 101K 1.5M 9K

View Details

fatwang2 @fatwang2ai

19 hours ago

Good night. 🌙 jev.s1.dev

fatwang2 @fatwang2ai

a day ago

Built Jev Search -- free and open source, powered by @typesafeai Jev and @search1api_dev search API. Ask in plain language. Jev picks where to search and ranks what comes back. Try it now: jev.s1.dev

7 25 272 148K 375

28 67 919 116K 990

View Details

Akshay 🚀 @akshay_pachaar

10 hours ago

x.com/i/article/2100…

20 105 766 56K 1K

View Details

Username To Be Determined @IgorSemshovv

21 hours ago

это я ставлю лайки на все посты про новый вид кошек

13 2K 21K 115K 143

View Details

Agita Christie @pagan_hoetry

13 hours ago

My cat has been waking me up at 5am daily and I can’t figure out what she wants (not interested in food). This morning I decided to just walk around with her to see what’s up and she led me to the office and stared at my office chair until I sat down. She wants me to work

308 5K 123K 1.6M 3K

View Details

George Proust @GeorgeProust

11 hours ago

As I said in early March. You can't lose 16 bases, with some 2,000 troops at each, and have only 13 dead. And how about the 82nd in Kuwait? Big losses.

barry with the NED @bonzerbarry

11 hours ago

WAPO: More U.S. troops have died amid Iran war than Pentagon has disclosed publicly. Officials who spoke with The Post expressed frustration that the additional loss of life in the Middle East is not reflected in the Defense Department’s casualty database.