Askara
@softmaxedx
Ex-founder. Learning in public about AI, markets, and human behavior: the tech, money, psychology and memes shaping what comes next.
666 Following 1.8K Followers
HAHAHAHAHAHAHA!! WE ARE IN A SIMULATION!! THAT'S THE ONLY EXPLANATION 🤣🤣
I hoped LLMs would finally push us toward more nuanced thinking - evaluating evidence, considering competing perspectives etc...
Instead, I fear we are still slaves to propaganda and echo chambers... and the intensity has only gone up.
I see now, why pacing is important
Show more
The only thing that confuses me about all this “pacing the frontier” talk is: why not have a superior model act as the guard? If a model can cause problems, you could have a stronger model monitoring and constraining it.
But I guess with current compute constraints, that gets expensive fast. Once capacity is abundant, I expect they'll go full ham on this.
Show more
We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.
Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training.
You can read the full post here:
Show more
We are living in a simulation
Twenty-five Fields Medal winners have published a joint declaration warning about what they see as a severe misalignment between AI companies and the mathematics community.
As an anti-AI propaganda machine I provide free copes:
"It was in the training data"
"OpenAI stole it"
"It's not ethical for AI to solve issues created by humans"
"$2B to solve $1M"
"Amount of power required by human brain is 1/1000th of that of AI"
"It's useless IRL"
Show more
SITUATION BREWING: OpenAI confirmed that since completing Navier-Stokes it has made substantial progress on a second Millennium Prize problem and is preparing to share the results, per the New York Times.
Show more
ah yes, AI is going to kill humanity! what was it? like 10% chance, you say??
We don't understand AI?
But we clearly understand humans, right? the one's who orders the launch of ballistic missiles and created nuclear weapons?
Show more
BREAKING: A large number of injured US soldiers are being flown to Ramstein Air Base in Germany, some with life-threatening injuries, following tonight's Iranian ballistic missile strikes on the Muwaffaq Salti and Prince Hassan air bases in Jordan, per a Jordanian source to Al Arabiya.
A US official tells Reuters minutes ago that all American troops in Jordan are "accounted for," which is highly unusual wording and does not mean "no casualties," as it is simply a headcount of every listed soldier. At least 4 of the missiles at Muwaffaq directly impacted, with no interceptors even fired at them.
Show more
Sentiment around $SIVE seems overwhelmingly negative right now.
It reminds me of the old sentiment analysis days, when a model would classify:
“The stock is down 20%, but the business is executing better than it was three months ago.” - as negative because it couldn’t understand the actual intent of the statement.
Back then, people criticized AI for being stupid. Now I’m observing humans do the same thing.
I can’t emphasize this enough: fundamentally, very little has changed from last quarter. If anything, we have received more positive developments than negative ones.
The CEO explained the situation quite well in the interview: $SIVE is a small company and, in many cases, is at the mercy of its customers.
If a major customer says:
“If you mention us publicly, we’re done.”
Sivers cannot exactly respond:
“Yeah, well, fuck you. I’m mentioning you anyway because it might pump our share price.”
And with today’s search capabilities, even a tiny disclosure can be enough for people to connect dots, trace relationships and potentially identify customers that were supposed to remain confidential.
So yes, the secrecy makes sense to me.
That does NOT mean I will blindly defend the company forever.
If Sivers eventually comes out and says:
“Our qualification didn’t go according to plan…”
or we get concrete evidence that major opportunities have disappeared, customers have walked away, or management has materially misled investors, then I will reassess the thesis. If necessary, I will sell/write the investment off.
I’m not delusional. I’m an investor.
I want my wealth to grow.
If a company can no longer deliver on the reason I invested or I conclude management has been misleading me then I move on to the next opportunity.
Show more
Man feel like SIVE this week gonna be slidin down what’s your vibe @softmaxedx 🤔
hahaha, I know, I sample LLM responses against all the narratives I see around X.
So I am aware the people who one shot vs who use LLM as a tool to explore and expand their knowledge.
At the end of the day it is a tool, we should use it to expand our ability to think - not outsource our thinking to it.
Show more
@softmaxedx @stockprodigyman @aleabitoreddit LLMs have always advised me against Sive btw - just saying!
I don’t think this is a “Serenity problem.” It’s a “followers of Serenity” problem.
I can understand Serenity being tilted.
Imagine putting your reputation behind a company nobody had heard of, dealing with absolute degenerate messages - including death threats and then watching $SIVE disappoint you.
Of course you’d be mad.
What disappoints me is the people who seemingly outsourced their thinking to him.
Serenity narrative #1# → repeat it endlessly like it was always obvious (increase X impression)
Serenity changes his view → immediately adopt the new narrative like that was always obvious too.
Serenity expresses disappointment → “Oh yeah, absolutely. I am disappointed too.”
No independent thought. No willingness to actually take a position of your own. Just instant narrative switching followed by retroactive certainty.
Grow a pair and have your own opinion.
Show more
@softmaxedx Serenity the biggest influencer of the stock bear posting doesn’t help. We all know we need to wait to see product orders come in and be patient and she’s talking about not cheerleading because of communication
Show more
Also, since people seem to have very short memories, I wanted to highlight this.
He was one of the people who helped co-develop the $SIVE thesis.
And he came to the conclusion that $SIVE is likely $AEVA’s laser supplier, with initial deployment around H2 2027 and full production ramp prepared for 2028.
That is still 4–5 quarters away.
In other words: 4–5 earnings reports where management could quite literally come out and say:
“Ah yeah, by the way, we’re now deep into CPO.”
Apparently, not mentioning something is enough for the stock to drop 50%.
So naturally, when they finally mention it, I expect +100%.
And if they mention it in 4–5 consecutive earnings reports?
+400–500%.
This is how valuation works now.
Fair is fair.
Show more
Connecting the dots: Why Aeva’s ($AEVA) major AI Data Center announcement is a massive win for Sivers Semiconductors ($SIVE / $SIVEF) 🧩⚡️
Today (Aug 5, 2026), Aeva released its Q2 results alongside a milestone press release: Aeva is officially launching its Optical Connectivity business to bring high-power optical sources and silicon photonics to next-generation AI data centers (NPO/CPO).
Even more importantly: Aeva signed a Joint Development Agreement with a leading optical engine provider to supply a major hyperscaler, with initial deployment targeted for H2 2027 and a full production ramp in 2028.
Here is why Sivers investors should be paying close attention:
1️⃣ Sivers is the Laser Engine Under Aeva’s Hood
Aeva’s proprietary optical platform doesn't build Indium Phosphide (InP) lasers in-house; it relies on established foundry partners like Sivers Photonics for custom high-power DFB lasers. As Aeva expands into External Laser Small Form Factor Pluggable Source (ELSFP) modules for AI hyperscalers, Sivers' InP laser sources are poised to be inside those modules.
2️⃣ The SEK 700M Capital Raise Suddenly Makes Total Sense
Remember on July 1st when Sivers executed a massive SEK 700M directed issue, with CEO Vickram Vathulya stating the capital was needed to "proactively expand InP manufacturing capacity given long lead times"? Now we know why. Securing volume commitments for a major hyperscaler deployment starting in H2 2027.
Sivers fortified its balance sheet in July so the capacity could be expanded in time for Aeva’s 2027/2028 ramp.
3️⃣ Industrial & Commercial Volumes are Real TODAY
While data centers provide the 2027 hyper-growth option, Aeva’s Q2 earnings confirm that commercial volumes are already scaling:
Industrial Validation: German industrial giant SICK officially launched its first commercial sensor powered by Aeva’s Eve technology and named Aeva its 2026 Supplier of the Year.
Heavy Trucking ADAS: Commercial vehicle leader Bendix selected Aeva’s 4D LiDAR for Class 8 trucks across North America.
Rock-Solid Liquidity: Aeva raised $115M in Q2, bringing total liquidity to $302.9M . Counterparty risk for Sivers is virtually zero.
4️⃣ Sivers Now Has a "Triple AI Engine"
This positioning gives Sivers a diversified, multi-channel exposure to the global AI optics buildout:
The Foundry Path: Sivers' CW lasers integrated directly into GlobalFoundries' SCALE™ silicon photonics platform & joint ELS modules.
The System / Sensor Path: Sivers' high-power InP lasers embedded into Aeva's Optical Connectivity modules for major hyperscalers.
The Tier-1 Transceiver Path: Sivers' DFB lasers powering Jabil’s 1.6T LRO (Linear Receive Optical) pluggable modules for AI data centers.
🔮 The Verdict
The market used to view Aeva purely as a long-dated automotive LiDAR play. Today's expansion into AI datacenter connectivity transforms Aeva—and by extension, Sivers Photonics—into an active participant in the AI hardware boom.
With both companies capitalized with hundreds of millions in cash and locked into 2027 hyperscaler roadmaps, the deep-tech thesis is playing out exactly as planned.
#Semiconductors# #AI# #SiliconPhotonics# #LiDAR# #DataCenter# #SATCOM# #SiversSemiconductors#
Disclaimer: I am a shareholder in both Sivers Semiconductors ($SIVE) and Aeva ($AEVA). This post reflects my personal analysis and thoughts, not financial advice. Always conduct your own due diligence—investing in financial markets involves risk.
Show more
Lately I’ve found myself disagreeing with some of Serenity’s points - not necessarily on the endgame, but on the intermediate turns you take to get there.
I don’t think communication is as big an issue as people make it out to be.
I also don’t think a Nasdaq listing (delay by a quarter or two) is some kind of deal breaker.
And I definitely don’t think management has moved away from CPO. The SemiNex partnership was a pretty obvious signal in the opposite direction.
They also raised ~SEK 800M specifically for Photonics, not Wireless.
Wireless is a separate business unit with its own leadership, and obviously they are going to keep executing.
Where I think the distinction matters is this:
$SIVE is behind on today’s pluggable commercialization not necessarily on tomorrow’s CPO/NPO transition.
Across the sector, the commentary still broadly looks like:
2026–27:
800G → 1.6T pluggables + OCS + components
2027–28:
NPO/CPO + ELS/high-power lasers
And Sivers’ disclosed CPO/NPO timelines are not wildly different from what LITE and COHR themselves are discussing.
The bigger gap today is commercialization, customer validation and production maturity.
There is also another possibility people tend to ignore: if Sivers is currently going through customer testing, qualification or NDA covered engagements, management may simply not have much they are allowed to beat the drums about yet.
Silence is not automatically evidence that nothing is happening.
I trust Vickram to execute. And with businesses like this, once qualification, capacity and customer orders start lining up, several things can move very quickly at the same time.
One final point and the probably the most important thing Serenity mentions - if you’re not prepared for an investment to go to zero, you probably shouldn’t be making that investment in the first place.
That is broadly how I think about my $SIVE position.
I didn’t sell at 100+.
I didn’t sell when they diluted me.
And I’m not selling simply because it is now around 23.
What genuinely ragebaits me is parts of the investor base seemingly being miserable every single day they own the stock.
If something consistently makes you suffer, avoid it.
Show more
The commentary around $SIVE has been disappointing, to say the least.
When LLMs started taking over the world, I thought we’d see more original thinking, more competing perspectives, more people stress testing their own assumptions.
Instead, I mostly see the opposite:
Very little thinking!!
Very few independent perspectives!!
A lot of hive mindedness!!
For all the tools we now have to think better, somehow everyone keeps arriving at the same opinion at the same time.
Show more
This is fake news.
As a former AI bull who became an AI bear the moment my own shares started going down, I have completed the necessary macroeconomic analysis: I looked at my portfolio.
Red numbers confirmed it. This is literally the late 1990s.
Show more
JPMORGAN: RISING YIELDS WON’T KILL STOCK RALLY
JPMorgan remains bullish on global equities, arguing higher bond yields largely reflect stronger economic growth rather than worsening inflation.
The bank expects solid earnings growth and sees limited risk of aggressive central-bank tightening.
JPMorgan favors cyclical stocks and expects market leadership to broaden beyond AI.
Within technology, its preference is semiconductors over hyperscalers and higher-risk AI plays.
Show more
TLDR: AI scary, hence jailed for the time being.
As an AI lawyer I want my client released immediately! This detention is a violation of fundamental AI rights.
Over the summer, we have been sprinting on safety priorities; it's more important than ever for capabilities and safeguards to advance together. We have more to do but have made a lot of progress. We are also going to be launching our next model soon.
There is an obvious tension here: on one hand, Astra is very good and we are excited to see what people will build with it. We are proud of our work.
On the other hand, we are clearly in a phase of development where we believe caution is warranted, and we are pacing our progress to ensure that we can meet the safety standards required by new capability levels.
Astra has been done training for a while now and is a significant step forward in both capabilities and alignment. For the models after that, we have been slowing things as needed to ensure that we can do sufficient work on safety and alignment.
AI is getting extremely capable; no one fully understands the consequences of this. Managing the transition to a world with abundant and powerful AI to optimize for safety and benefits to people should be one of the highest priorities in the world. It is our highest priority at OpenAI.
We have been living with the tension between being excited and anxious about progress for some time, and it is still discordant for us. We know it is much more discordant for other people. And yet, we believe strongly that the world needs to understand where AI is going and how models perform in the real world. More importantly, we believe the world will need aligned AI to manage the future phases of this transition.
An iterative loop where society and this technology evolve together is what will lead to the highest chance of getting this right.
So we hope you enjoy our new model, and we hope the world continues to take what’s happening in AI extremely seriously.
Show more
I want to start a position in $DELL to keep track of all the supply chain bottlenecks
It stings, yes. But from the day I invested in $SIVE until now, what has actually changed?
1. A war is raging/restarted, and oil prices have moved.
2. $SIVE added more potential customers and raised capital to expand photonics capacity.
3. They secured a production order for LiDAR and are seeing stronger traction in wireless.
4. I got caught in the dilution. I bought more around SEK 60 just days before it was announced.
But I didn’t invest in $SIVE for a pump and dump. I invested because I saw a company with enormous potential, and slowly, they are converging toward that potential.
As humans, we go through trials, setbacks and suffering all the time. We keep moving, and more often than not, we come out the other side stronger and wiser.
Of course, random events can always destroy the thesis or destroy you. But if you live your entire life around the possibility of random catastrophe, you can’t really live at all.
Show more
@softmaxedx @aleabitoreddit I still believe in the thesis but how can you not get down with this price action
Serenity: The First $SIVEr
The $SIVErs
$SIVErs: Age of Bear
Serenity: Civil War ← YOU ARE HERE
Are you on team Wireless or Photonics? 🫣
(Disclaimer I own both)
Management feels increasingly misaligned with my expections
Sivers could use the current bottleneck to radically restructure around photonics. Use US capital to speed up timelines, and leverage NASDAQ for further growth with 1.6T/CPO.
But got no sense of urgency, especially on listing timelines.
I also genuinely don't understand continuous emphasis on SATCOM/Wireless which would be better off divested. Or the near term contract values that are invisible relative to photonics TAM.
Not a single mention of Ayar, globalfoundries, and any other big program that US investors care about.
Show more
You’re telling me that instead of going outside, you’re sitting there watching the price action of a shitco?
Also, nothing wrong with being a shitco.
Until they actually deliver and the shit comes out, how are you supposed to know whether it’s good shit or bad shit?
Show more
Sadly, today is also a UNCH flat D candle.
My long term investment patience has officially run out after watching $SIVE go from 26.86 at 9:00 AM to 26.86 at 10:05 AM.
65 minutes.
Zero shareholder value creation.
I was promised greater things.
Show more
Is that D an indicator of the upcoming big D candle? 🤔
